6s Manual v1.0

6S User Guide Version 1, November 3, 1995
1
Second Simulation of the Satellite Signal
in the Solar Spectrum (6S)
AIR FRANCE
- KRAPS -
FANS CLUB
Vermote E.
(1,* )
, Tanr D.
(2)
, Deuz J. L.
(2)
,
Herman M.
(2)
, and Morcrette J. J.
(3)
(1) Department of Geography, University of Maryland
address for correspondence:
NASA-Goddard Space Flight Center-Code 923
Greenbelt, MD 20771, USA
(2) Laboratoire d'Optique Atmosphrique, URA CNRS 713
Universit des Sciences et Technologies de Lille
59655 Villeneuve d'Ascq Cdex, France
(3) European Centre for Medium Range Weather Forecast
Shinfield Park, Reading
Berkshire RG2 9AX, United Kingdom

*
Formerly affiliated to Laboratoire d'Optique Atmosphrique
2
3
Contents
ABSTRACT 5
INTRODUCTION 7
DESCRIPTION OF ATMOSPHERIC EFFECTS IN SATELLITE OBSERVATIONS 11
1- Absorbing effects 11
2- Scattering effects 13
2.1 Case of a Lambertian homogenous target 13
2.2 Environment function 15
2.3 Intrinsic atmospheric reflectance 17
2.4 Airplane and elevated target simulations 19
2.5 Directional effect of the target 22
2.6 Atmospheric correction scheme 24
3- Interaction between absorption and scattering effects 24
CONCLUSION 27
FIGURES 28
ACKNOWLEDGMENTS 33
APPENDIX:
I Description of the computer code 35
II Examples of inputs and outputs 53
III Description of the subroutines 57
4
5
ABSTRACT
The remote sensing from satellite or airborne platform of land or sea surface in the visible
and near infrared is strongly affected by the presence of the atmosphere on the path Sun-target-
Sensor. The following manual presents 6S (Second Simulation of the Satellite Signal in the Solar
Spectrum), a code enabling simulation of the above problem. The 6S code is the improved version
of 5S developed by the Laboratoire d'Optique Atmospherique, 10 years ago. It enables to simulate
plane observations, to account for on elevated targets, to take into account for non lambertian
surface boundary conditions, and new gases (CH
4
,N
2
O,CO) are now integrated in the computation
of the gaseous transmission. The computation accuracy regarding Rayleigh and aerosols scattering
effects has been improved by the use of state of the art approximation and use of the successive
order of scattering method. The step used for spectral integration has been improved to 2.5
nanometers. Finally, all community approved results (e.g.: computation of Rayleigh optical depth )
have been integrated in the code.
6
7
INTRODUCTION
In the solar spectrum, sensors on meteorological or Earth remote sensing satellites measure
the radiance reflected by the atmosphere-Earth surface system illuminated by the sun. This signal
depends on the surface reflectance, but it is also perturbed by two atmospheric processes, the
gaseous absorption and the scattering by molecules and aerosols.
In the ideal case (without interfering atmosphere), the solar radiation illuminates the surface.
A fraction of the incoming photons is absorbed by the surface, whereas the remaining photons are
reflected back to space. Therefore, the measured radiance directly depends on the surface
properties: this radiance is the useful signal as much as it characterizes the actual surface
reflectance.
In the actual case, the signal is perturbed by the atmosphere. Only a fraction of the photons
coming from the target reaches the satellite sensor, typically 80% at 0.85 m and 50% at 0.45 m, so
that the target seems less reflecting. The missing photons have been lost through two processes:
absorption and scattering.
Some photons have been absorbed by aerosols or atmospheric gases (principally O
3
, H
2
O, O
2
,
CO
2
, CH
4
and N
2
O). Generally, absorption by aerosols is small, and satellite sensor channels avoid
the molecular absorption bands. Thus, in this study, the absorption effect is a correction factor.
Nevertheless, we have corrected for the absorption as accurately as possible which allows this code
8
to be used to help to define new satellite sensor spectral channel. The absorption is computed using
statistical band models with a 10cm
-1
resolution.
While some photons are absorbed others are scattered. The interaction with molecules or non-
absorbing aerosols is elastic, and the photons are immediately re-emitted in a direction other than
the incident one. After one or several such scattering processes, these photons leave the atmosphere
and must be counted in the budget of photons reaching the satellite sensor. However, their paths are
more complex than the previous direct paths as explained hereafter.
Firstly, let us consider the photons coming from the sun that do not reach the surface and are
backscattered toward the space. They take part in our radiative balance. This signal is an
interference term; it does not carry any information about the target.
9
On the sun-surface path, the remaining photons contribute to the illumination of the ground by
the way of scattered paths and compensate the attenuation of the direct solar paths.
This diffuse component has therefore to be considered in the useful signal.
In a second step, let us consider the photons scattered by the atmosphere on the surface-
satellite path. By the same way, a fraction will be scattered toward the sensor.
This component has to be carefully considered. If the surface is uniform, it is an useful
component but if the surface has a patchy structure, this term will introduce environment effects
that will be a perturbation.
Finally we have to consider the photons which are backscattered by the atmosphere to the
surface and that create a third component of its illumination; it is the trapping effect, the photons
10
successively interact with the surface and the atmosphere but, generally, the convergence is fast and
after one or two interactions, the phenomenon can be neglected.
From this qualitative description, we can then infer a modelization of the atmospheric effects
based upon:
an accurate approach of the absorption by atmospheric gases
a complete treatment of the scattering processes
and an approximation for the interaction between the two processes.
11
DESCRIPTION OF THE ATMOSPHERIC EFFECTS IN SATELLITE OBSERVATIONS
1 Absorbing effects
In the solar spectrum the atmospheric gaseous absorption is principally due to:
oxygen (O
2
);
ozone (O
3
);
water vapor (H
2
O);
carbon dioxide (CO
2
);
methane (CH
4
);
nitrous oxide (N
2
O).
O
2,
CO
2
, CH
4
, and N
2
O are assumed constant and uniformly mixed in the atmosphere, H
2
O and O
3
concentrations depend on the time and the location. The latter two are the most important gases in
our study.
Gases absorb the radiation by changes of rotational, vibrational or electronic states. The
variations of rotational energy are weak and correspond to the emission or to the absorption of
photons of weak frequency, which are then located in microwave or far-infrared range. The
vibrational transitions correspond to greater energy which open to absorption spectrum in the near
infrared. They can also take place with rotational transitions and then give rise to vibrational-
rotational bands. Lastly, electronic transitions correspond to more important energy and give rise to
absorption or emission bands in the visible and the ultra-violet range. Since these transitions occur
at discrete values, the absorption coefficients vary very quickly with the frequency and present a
very complex structure.
Once the position, intensity and shape of each line in an absorption band is known, the
absorption can be then exactly computed using line-by-line integrations. Such a model requires a
very large computational time which makes it necessary to use an equivalent band model. We have
divided the solar spectrum into spectral intervals of 10cm
-1
width using HITRAN database. This
width allows a good description of the spectral variations of the transmission and of the overlapping
between absorption bands of different gases. Moreover, this allows enough absorption lines to be
contained in order that the transmission models remain well adapted to the problem.
12
We chose two random exponential band models, Goody (1964) for H
2
O and Malkmus (1967)
for the other gases. When several gases present absorption bands, we put the total transmission as
the product of each gas transmission.
A complete description of the statistical models is given in Appendix III (subroutine
ABSTRA), Here we present the results for each gas between 0.25 m and 4 m for a standard
atmosphere model corresponding to the US-1962 standard atmosphere and a sun at the nadir (air
mass is 1.0). These resuts are depicted in Figures I-1 to I-6
The H
2
O contribution (Fig. I-1) affects mainly wavelengths greater than 0.7 m. On the other
hand, O
3
presents a significant absorption between 0.55 and 0.65 m and limits the earth
observations at wavelengths less than 0.35 m (Fig. I-2). The CO
2
contribution occurs beyond 1 m,
but more weakly than H
2
O and only perturbing the water vapor windows (Fig. I-3). The O
2
influence is limited to a very strong band about 0.7 m (Fig. I-4). The CH
4
presents two absorption
bands at 2.3 and 3.35 m (Fig. I-5). Finally, the N
2
O contribution appears in two bands at 2.9 and
3.9 m (Fig. I-6).
In summary, we have, in the Solar Spectrum, good atmospheric windows for terrestrial
observations by satellites:
in the visible: between 0.40 and 0.75 m
in the near and middle infrared: at about 0.85, 1.06, 1.22, 1.60, 2.20 m.
Fig. I-7 to I-14 display these windows
13
2 Scattering effects
2.1 Case of a Lambertian uniform target
In a first step, assuming that the surface is of uniform Lambertian reflectance and the
atmosphere horizontally uniform and various, the measured quantities will be expressed in terms of
equivalent reflectance, * defined as:
*
L
s
E
s
(1)
with L the measured radiance, E
s
the solar flux at the top of the atmosphere and
s
=cos(
s
)
where
s
is the sun zenith angle.
The optical thickness of the atmosphere will be noted , the viewing direction will be
referenced by the zenith angle
V
and the azimuth angle
V
, and the sun angles by
s
and
s
.
t
will
be the reflectance of the target. Absorption problems will not be considered in this part.
s
v
n
s
-
v
It is convenient to express the signal received by the satellite as a function of successive
orders of radiation interactions in the coupled surface-atmosphere system.
For the surface illumination, we have by decreasing magnitudes:
the direct solar flux attenuated by the atmosphere
/
s e
14
the scattered flux on the first sun-surface path; it is independent from surface
properties and will be noted by a diffuse transmittance factor t
d
(
s
) defined as
t
d
(
s
)
E
sol
diff
(
s
)
s
E
s
where E
sol
diff
represents the downward diffuse solar irradiance
a second scattered flux due to the trapping mechanism; it is depending on the
environment of the target and corresponds to the successive reflections and scattering
between the surface and the atmosphere. If the spherical albedo of the atmosphere is
noted S, we can write this terms as
e
/
s
t
d
(
s
)
t
S
t
2
S
2
...
The total normalized illumination at the surface level is then written
T(
s
) / [1
t
S] (2)
where T(
s
) is the total transmittance:
T(
s
) e
/
s
t
d
(
s
) (3)
At the satellite level, the radiance results from:
the contribution of the total (direct plus diffuse) solar radiation reflected by the surface
and directly transmitted from the surface to the sensor expressed by e
- /
v with
v
=
cos(
v
)

15
the intrinsic atmospheric radiance expressed in terms of reflectances by function
a
(
s
,
v
,
s
-
V
)
the contribution of the environment which reflects the total (direct + diffuse) flux, the
photons reaching the sensor by scattering; we note this new atmospheric diffuse
transmittance t
d
'
(
v
)
So, the apparent reflectance * at the satellite level can be expressed as
*(
s
,
v
,
s v
)
a
(
s
,
v
,
s v
)
T(
s
)
1
t
S
(
t
e
/
v
t
t
d
'
(
v
)) (4)
In actual fact, the functions t
d
(
s
) and t'
d
(
v
) are identical in accordance with the reciprocity
principle and * can be rewritten
*(
s
,
v
,
s v
)
a
(
s
,
v
,
s v
)
t
1
t
S
T(
s
)T(
v
) (5)
with
T(
v
) e
/
v
t
d
(
v
) (6)
2.2 Environment function
Let us assume that the surface reflectance is now not uniform. In a first approach, we shall
consider a small target M of reflectance
c
(M) with an uniform environment of reflectance
e
(M). In
this case, Eq.(4) remains interesting by its formalism, it gives the weight of each reflectance and
*
can be now written as
*(
s
,
v
,
s v
)
a
(
s
,
v
,
s v
)
T(
s
)
1
e
S
(
c
(M)e
/
v
e
(M)t
d
(
v
)) (7)
16
If we now consider a patchy structure, we keep the same formalism as in Eq.(7), we only
have to define a new environment reflectance noted < (M)> which represents a spatial average of
each pixel reflectance over the whole surface.
*(
s
,
v
,
s v
)
a
(
s
,
v
,
s v
)
T(
s
)
1 (M) S
(
c
(M)e
/
v
(M) t
d
(
v
)) (8)
This spatial average must be weighted by an atmospheric function which takes into account
the efficiency of a point M' according to the distance from the point M, so < (M)> will be given by
(M)
1
t
d
(
v
)
' (x,y) e(x,y,
v
)dx dy (9)
M(0,0)
M'(x,y)
x
y
where
M is the origin of the x, y coordinate system;
'(x,y) is the reflectance of the point M'(x,y);
e(x,y,
V
) is the contribution to the diffuse transmission t
d
(
v
) per unit area of an isotropic
unit source placed at M'(x,y).
The use of the same average reflectance < (M)> for the high orders of interaction (that is
the geometrical series) is incorrect but reasonable because in most cases the contribution from these
terms does not exceed 10-15% of the contribution of the first order term.
If we restrict the modeling to the case of vertical observations (or nearly,
v
<30), Eq.(9)
can be written
17
(M) '(r, )rp(r)drd
0 0
2
(10)
with
r p(r)
re(r, ,0)
t
d
(0)
(11)
where (r, ) are the polar coordinates of a given surface point with object pixel M at the origin. For
off-nadir observation, we will see in th subroutine ENVIRO how to handle the problem.
Actually, it is impossible to solve this problem exactly, using Eq. (10) because of the
computational time involved. Generally, we can define a characteristic size r of the target whose
reflectance can be assumed uniform. By the same way, we can compute the environment reflectance
e
reflected by a simple arithmetical mean and we estimate < (M)> in the following manner.
Considers a circular target of radius r and reflectance
c
surrounded by homogeneous surface of
reflectance
e.
Eq.(10) becomes:
(M)
c
F(r) (1 F(r))
e
(12)
with
F(r) 2 r'p(r')dr'
0
r
(13)
which gives the relative contribution to < (M)> of surface points not farther than a distance r apart
from the origin.
2.3 Intrinsic atmospheric reflectance
The intrinsic atmospheric reflectance observed over black target,
r
+
a
is written here as the
sum of aerosols and Rayleigh contributions. This decomposition is not valid at short wavelengths
(less than 0.45 m) or at large sun and view zenith angles.
18
2.3.1 Rayleigh
Atmospheric reflectance
For isotropic scattering, Chandrasekhar (1960) showed how solutions derived for small
optical thicknesses may be extended to larger values of . He expressed the atmospheric reflectance
a
(
s v, v- s
as
a
(
s
,
v
,
v s
)
a
1
(
s
,
v
,
v s
) (1 e
/
s
)(1 e
/
v
) ( ) (14)
where
a
1
(
s
,
v
,
v s
) is the single-scattering contribution and the second term accounts
roughly for higher orders of scattering.
In 6S, we used this approach to compute the molecular scattering reflectance.
Transmission function
The transmission function refers to the normalized flux measured at the surface. There are
several approximate expressions based on the two-stream methods for computing the transmitted
flux. The accuracy of these expressions depends on the scattering properties of the atmospheric
layer (thick or thin clouds or aerosols) and on the geometrical conditions. The delta-Eddington
method, which has been proved to be well suited for our conditions, has been selected. Since
molecular scattering is conservative (
0
=1) and the anisotropy factor g is equal to zero, we may
write
T( )
[2 / 3 ] [2/ 3 ]e
/
4 / 3
(15)
where is the cosine of the solar and/or observational zenith angle and is the optical thickness.
Spherical albedo
In conservative cases such as molecular scattering, the spherical albedo s is given by
s T( ) d
0
1
(16)
where T( ) has been given in Eq.(15). Using Eqs.(15) and (16), the spherical albedo can be written as
s
1
4 3
[ E
3
( ) 6E
4
( )] (17)
where E
3
( ) and E
4
( ) are exponential integrals for the argument .
19
2.3.2 aerosol
For the aerosols, the optical scattering parameters were computed using pre-defined models,
Continental, Maritime, Urban or user's model based on a mixture of 4 basic components, Dust-Like,
Oceanic, Water-Soluble, and Soot (provided by the World Climate Program). In 6S, these options
are still usable but we added 2 news models, the Background Desertic Model and the Stratospheric
model, and we also offer to the user the possibility of making up his own aerosol model (4
components maximum, see the description of the subroutine MIE for details).
In 5S, the scattering properties pertaining to the aerosol layer was performed using the
Sobolev (1975) approximation for the reflectance, Zdunkowsky et al. (1980) for the transmittance
and a semi-empirical formula for the spherical albedo. The reason was to provide the user having
limited computing ressources with a fast approximation. The drawbacks to using these
approximations were that the accuracy of the computations could be off by a few percent in
reflectance units especially at large view and sun angles or high optical thicknesses. In addition,
these approximations could be completely insufficient to handle the integration of the downward
radiance field with non-lambertian ground conditions, a problem in simulating BRDF. The new
scheme used to compute the aerosol+rayleigh coupled system relies on the Sucessive Order of
Scattering method. The accuracy of such a scheme is better than 1 10
-4
in reflectance units. Other
improvements on the 5S model include an atmosphere divided into 13 layers which enables exact
simulations of airborne observations; the downward radiation field is computed for a quadrature of
13 gaussian emerging angles which will provide the necessary inputs for BRDF simulations (see
2.5).
2.4 Airplane and elevated target simulations
2.4.1 Elevated target simulation
For a target not at sea level, Eq.(5) is modified as follows:
*
(
s
,
v
,
s v
, z
t
) Tg(
s
,
v
,z
t
) [
r
(z
t
)
a
t
1 S(z
t
)
t
T(
v
,z
t
)T(
s
,z
t
)] (18)
The target altitude or pressure indicates the amount of scatterers above the target (molecules
and aerosols) and the amount of gaseous absorbants. The target altitude z
t
is handled in the
following manner: After the atmospheric profile and the target altitude or pressure are selected, a
new atmospheric profile is computed by stripping out the atmospheric level above target altitude
and interpolating if necessary. This way, an exact computation of the atmospheric parameters are
computed, without any kind of approximation which will account for coupled pressure-temperature
effect on absorption.
20
Gaseous absorption
In most cases, only the integrated content may be modified. The user still has the option to
enter the total amount of H
2
O and O
3
, but in this case the quantity entered must be representative
of the level measured or estimated at the target location. The influence of target altitude on Tg has
been evaluated. The absorption effect by O
3
is not sensitive to target altitude, because the ozone
layer is located in the upper levels of the atmosphere, the variation of the altitude of the target does
not modify this effect. The target's altitude has an important effect on the absorption by H
2
O
because most of the water vapor is located in the lower atmosphere. However the exact sensitivity
of the target's altitude on water vapor absorption cannot be generalized because water vapor profile
is variable.
Scattering function
The effect of target altitude on molecular optical thickness is accounted for in 6S. A good
approximation, if needed, is to consider that
r
is proportionnal to the pressure at target level. As the
amount and types of aerosols are entered as parameters, the aerosol characteristics implicity depend
on target altitude because measured at target location.
2.4.2 Airborne sensor simulation
In case of a sensor inside the atmosphere (aircraft simulation), Eq (5) is modified as
following:
*
(
s
,
v
,
s v
, z) Tg(
s
,
v
,z) [
r
(z)
a
(z)
t
1 S(z)
t
T(
v
, z)T(
s
,z)] (19)
Gaseous absorption
Gaseous absorption is computed with a technique similar to the one used in the case of a
target not at sea level. Only, the upward path is modified: the atmosphere level above the sensor
altitude are stripped, so the computation is done only to the altitude of the sensor (interpolation of
the atmospheric profile is conducted if necessary). The effect of altitude on gaseous transmission
has been computed. For visible wavelength ranges, O
3
absorption on the target-sensor path is no
longer accounted for due to the fact these molecules are located above most of aircraft sensors. H
2
O
absorption is very dependent on sensor altitude up to 4km. If the observed channel is sensitive to
water vapor absorption (as it is in the case of AVHRR channel 2) we recommend that additionnal
measurements of water vapor should be taken from the aircraft. An additional option has been set
21
up in 6S for this purpose and enables the user to enter ozone and water vapor contents (as well as
aerosol content) for the portion of the atmosphere located under the plane.
atmospheric reflectance and transmittance
In most cases the simple approximation "equivalent atmosphere", for atmospheric
reflectance and transmittance is sufficiently accurate, ie:
r
(z)
r
(z ,
r
(0 z)) (20)
T(
r
,
v
,z) T(
r
(0 z),
v
, z ) (21)
which means that computations are performed for optical thicknesses corresponding to the
integration between the surface and the sensors.
In 6S, the computation is performed exactly by defining one of the multiple layers used in
the Sucessive Order of Scattering as the alitude of the sensor. This enables exact computation of
both reflectance and transmission term for a realistic mixing between aerosol and Rayleigh.
2.4.3 Non uniform target
In 5S, the case of a non-homogeneous target is solved using Eqs.(8) and (12) where the
function F(r) is defined as
F(r)
t
d
r
(
v
)F
r
(r) t
d
a
(
v
)F
a
(r)
t
d
(
v
)
(22)
The problem of a target not at sea level, as long as we consider that the target and
environment are at the same altitude, can be solved just by modifying the rayleigh optical thickness.
In the case of aircraft observations, first we have to take into account the reduction of the
amount of scatterers under the plane. This can be done just by adjusting t
d
r
(
v
) and t
d
a
(
v
) to
t
d
r
(z,
v
) and t
d
a
(z,
v
). Once this has been done, the principal part of the effect accounted for is a
global reduction (factor 5-10 for an altitude of flight of 6 km) of the "environment effect". The
second effect is the dependence of F
r
(r) and F
a
(r) upon altitude of the sensor. Monte Carlo
simulation of F
r
(r) and F
a
(r) have been performed for different altitudes of the sensor (0.5,....12km)
22
and included in 6S as a database. In case of a plane observation, the closest simulated altitudes are
used to interpolate the environment function at the aircraft altitude.
2.5 Directional effect of the target
2.5.1. BRDF
In 6S, the coupling of BRDF with the atmosphere is addressed, the contribution of the target
to the signal at the top of the atmosphere is decomposed as the sum of four terms:
the photons directly transmitted from the sun to the target and directly reflected back to the
sensor
e
/
s
t
(
s
,
v
, )e
/
v
(23)
s v
=
s
-
v
the photons scattered by the atmosphere then reflected by the target and directly transmitted
to the sensor,
t
d
(
s
)
t
(
s
,
v
, )
/
v e t
d
(
s
)
L (
s
, , ' )
0
1
0
2
t
( ,
v
, ' ) d d '
L (
s
, , ')
0
1
0
2
d d '
/
v e (24)
s
v

'
s
the photons directly transmited to the target but scattered by the atmosphere on their way to
the sensor,
t
d
(
v
)
t
'
(
s
,
v
, ) e
/
s
t
d
(
v
)
t
(
v
,
s
, ) e
/
s
(25)
23
s
v
'

'
v
the photons having at least two interactions with the atmosphere.
t
d
(
v
) t
d
(
s
)
t
t
d
(
v
) t
d
(
s
)
t
'
(
s
,
v
, ) (26)
s
'
'
v
In 6S, the first three contributions are computed exactly using the downward radiation field
as given by the Sucessive Order of Scattering method. The contribution which involves at least two
interactions between the atmosphere and the BRDF (Eq.(26)) is approximated by taking
t
equal to
the the hemispherical albedo of the target
t s
0
1
t
(
s
,
0
2
v
,
0
1
' )d
v
d d
s
(27)
This approximation is required because the exact computation needs a double integration
which would result in very long computation times. It is in addition justified by the limited impact
on the total signal of this last contribution mitigated by t
d
(
s
) t
d
(
v
) and also because the radiation
field after one interaction trends to be isotropic.
24
2.5.2. Sunglint
The solar radiation directly reflected by the water surface is computed exactly with the
Snell-Fresnel laws. For a rough sea surface, the reflection is conditioned by the wind and its
description is by necessity a statistical one. Thus, the model surface can be computed numerically
by many facets whose slopes are described by a Gaussian distribution (Cox and Munk, 1954). In
6S, this slope distribution is considered anisotropic (depending upon wind direction). Operationally,
the scheme to compute the reflection by the sunglint is identical to BRDF ones.
2.6 Atmospheric correction scheme.
An input parameter allows to activate atmospheric correction mode. In this case, the ground
is considered to be Lambertian , and as the atmospheric condition are known the code retrieves the
atmospherically corrected reflectance value
ac
that will produce the reflectance (or the radiance)
equal to the apparent reflectance
i
*
(Eq.(1)) (or apparent radiance) entered as input. Following Eq.
5,
ac
be determined as:
ac
(
s, v
,
s v
)
ac
'
1
ac
'
s
(28)
with
ac
'
i
*
(
s
,
v
,
s v
)
Tg
a
(
s
,
v
,
s v
)
T(
s
) T(
v
)
(29)
where
i
*
is the input parameter and where all the others parameters are computed in the
atmospheric conditions described by the user.
3 Interaction between absorption and scattering effects
The atmospheric radiation depends on atmospheric transmittances which can be modified by
the contribution of absorbing gases. The gaseous absorption must therefore be computed for each
scattering path. We can simplify the problem and separate the two processes.
Let us consider the O
3
gas. It is localized at altitudes where molecules and aerosols are
sparse, then the incident and reflected radiation go through the O
3
layer nearly without scattering.
25
So, we compute the atmospheric radiation * without accounting for O
3.
We modify the signal by
the only direct transmittance which corresponds to the twofold paths.
Tg
O
3
(
s
,
v
,U
O
3
) (30)
where U
O
3
is the absorber amount.
In the case of H
2
O and CO
2
, the absorption bands occur in the near and middle infrared. In
this range, the molecular scattering becomes negligible and we only have to take into account the
aerosol scattering. Some simulations, made for realistic atmospheres, show that beyond 0.850m,
the 1st and 2nd scattering constitute almost the whole diffuse radiation. Because the aerosol phase
function has a large forward scattering, the diffuse paths are not completely separate from direct
paths and we shall consider that the influence of CO
2
or H
2
O upon scattered flux is the same as for
direct solar flux. So the diffuse and transmitted flux will be both reduced by the factor
Tg
H
2
O
(
s
,
v
,U
H
2
O
)Tg
CO
2
(
s
,
v
,U
CO
2
) (31)
For O
2
, the absorption is very localized and presents a very narrow band, so we take into
account this gas by the same way:
Tg
O
2
(
s
,
v
,U
O
2
) (32)
We have checked our approximation from a simulation in the visible Meteosat spectral
band. This band is large enough (0.35-1.10 m) to include both scattering effects and water vapor
absorption. To amplify the coupling effect, we have taken a tropical atmosphere model (H
2
O
content is about 4g/cm
2
) and a continental aerosol model with large optical depth (
0.55 m
=1). For
=20cm
-1
, we have:
s
T
exact
(
S
) T
est
(
s
)
1.0 0.837 0.843
0.5 0.692 0.693
0.1 0.401 0.363
Estimation of the accuracy of our approximation
The comparison between exact computations and results obtained from our approach is
reported in the above table for three solar angles. Except for a grazing incidence (
s
=0.1), we have a
good agreement and the error related to this approximation is smaller than one percent.
26
For atmospheric reflectance, the problem is slightly different. The diffuse paths are not close
to the direct paths. The photons are scattered at large angles away from the forward direction. For
O
3
, the same approximation is still justified because it is based upon the weak interaction between
the two processes. For H
2
O, the scale height is comparable to the aerosols scale height while CO
2
is
uniformly mixed. Therefore, for these two gases our approximation is less valid but remains
sufficiently accurate as long as the gaseous absorption is small, i.e. in the atmospheric windows.
For getting the maximum uncertainties resulting from the approximation, we consider 3
extreme cases in 6S: the water vapor above the aerosol layer (maximum absorption, see (33-a) for
i=3), the water vapor under the aerosol layer (minimum absorption, see (33-a) for i=1), and an
average case where we consider that half of the water vapor present in the atmosphere absorbs the
aerosol path radiance (see (33-a) for i=2). In that respect, the satellite signal is written
TOA
i 1,3
(
s
,
v
,
s v
) Tg
OG
(
s
,
v
) [
R A
(
R A R
)Tg
H
2
O
(
s
,
v
,
i 1
2
U
H
2
O
)
T (
s
) T (
v
)
s
1 S
s
Tg
H
2
O
(
s
,
v
,U
H
2
O
) ] (33-a)
where Tg
OG
refers to the gaseous transmission for other gases than water vapor, Tg
H2O
refers to
H
2
O absorption and
R+A
-
R
is an estimate of the aerosol intrinsic reflectance. For each cases, we
compute the top of the atmosphere reflectance, so the uncertainty due to the vertical distribtution of
aerosol versus water vapor can be considered.Tg
OG
is computed on the sun-target-satellite direct
paths using
Tg
OG
(
s
,
v
) Tg
i
(
s
,
i 1
8
v
,U
i
) (33-b)
27
CONCLUSION
An accurate analytical expression of the reflectance measured by a satellite-sensor or by a sensor
aboard an aircraft has been established.
In the case of Lambertian and uniform surface, we completely separate the surface-
atmosphere system and define three atmospheric functions.
- the intrinsic atmospheric signal component
- the total transmittance
- the spherical albedo.
The coupled BRDF-atmosphere are accounted for according to the same scheme by
definition of mean angular reflectances which are depending on atmospheric properties.
In the case of a non-uniform ground, we have defined a spatial average reflectance which
remains slightly dependent on atmospheric properties by the definition of the function F(r).
This function allows us to take into account the main blurring atmospheric effect for non-
uniform surfaces.
In both cases, it is possible to consider elevated targets.
The absorption by atmospheric gases is computed separately as a simple multiplicative factor. The
solar spectrum is divided into spectral intervals of 10cm
-1
width.
In addition, a scheme of atmospheric correction is available.
The computational aspect of our work is now detailed in the following appendix:
Appendix I : General description of the computer code
Appendix II : Example of inputs and outputs
Appendix III : Complete description of the subroutines
28
0.0
0.2
0.4
0.6
0.8
1.0
0
.
2
0
.
4
0
.
6
0
.
8
1
.
0
1
.
2
1
.
4
1
.
6
1
.
8
2
.
0
2
.
2
2
.
4
2
.
6
2
.
8
3
.
0
3
.
2
3
.
4
3
.
6
3
.
8
4
.
0
Fig. I-1: Spectral transmittance of H
2
O
0.0
0.2
0.4
0.6
0.8
1.0
0
.
2
0
.
4
0
.
6
0
.
8
1
.
0
1
.
2
1
.
4
1
.
6
1
.
8
2
.
0
2
.
2
2
.
4
2
.
6
2
.
8
3
.
0
3
.
2
3
.
4
3
.
6
3
.
8
4
.
0
Fig. I-2: Spectral transmittance of O
3
29
0.0
0.2
0.4
0.6
0.8
1.0
0
.
2
0
.
4
0
.
6
0
.
8
1
.
0
1
.
2
1
.
4
1
.
6
1
.
8
2
.
0
2
.
2
2
.
4
2
.
6
2
.
8
3
.
0
3
.
2
3
.
4
3
.
6
3
.
8
4
.
0
Fig. I-3: Spectral transmittance of CO
2
0.0
0.2
0.4
0.6
0.8
1.0
0
.
2
0
.
4
0
.
6
0
.
8
1
.
0
1
.
2
1
.
4
1
.
6
1
.
8
2
.
0
2
.
2
2
.
4
2
.
6
2
.
8
3
.
0
3
.
2
3
.
4
3
.
6
3
.
8
4
.
0
Fig. I-4: Spectral transmittance of O
2
30
0.0
0.2
0.4
0.6
0.8
1.0
0
.
2
0
.
4
0
.
6
0
.
8
1
.
0
1
.
2
1
.
4
1
.
6
1
.
8
2
.
0
2
.
2
2
.
4
2
.
6
2
.
8
3
.
0
3
.
2
3
.
4
3
.
6
3
.
8
4
.
0
Fig. I-5: Spectral transmittance of N
2
O
0.0
0.2
0.4
0.6
0.8
1.0
0
.
2
0
.
4
0
.
6
0
.
8
1
.
0
1
.
2
1
.
4
1
.
6
1
.
8
2
.
0
2
.
2
2
.
4
2
.
6
2
.
8
3
.
0
3
.
2
3
.
4
3
.
6
3
.
8
4
.
0
Fig. I-6: Spectral transmittance of CH
4
31
0. 0
0. 2
0. 4
0. 6
0. 8
1. 0
0
.
3
0
.
4
0
.
5
0
.
6
0. 0
0. 2
0. 4
0. 6
0. 8
1. 0
0
.
6
0
.
7
0
.
8
Fig. I-7: Atmospheric window at 0.40 m Fig. I-8: Atmospheric window at 0.75 m
0. 0
0. 2
0. 4
0. 6
0. 8
1. 0
0
.
7
0
.
8
0
.
9
1
.
0
0. 0
0. 2
0. 4
0. 6
0. 8
1. 0
0
.
9
1
.
0
1
.
1
1
.
2
0. 0
0. 2
0. 4
0. 6
0. 8
1. 0
1
.
1
1
.
2
1
.
3
1
.
4
0. 0
0. 2
0. 4
0. 6
0. 8
1. 0
1
.
4
1
.
5
1
.
6
1
.
7
1
.
8
1
.
9
0. 0
0. 2
0. 4
0. 6
0. 8
1. 0
1
.
8
1
.
9
2
.
0
2
.
1
2
.
2
2
.
3
2
.
4
2
.
5
2
.
6
0. 0
0. 2
0. 4
0. 6
0. 8
1. 0
2
.
8
3
.
0
3
.
2
3
.
4
3
.
6
3
.
8
4
.
0
32
33
ACKNOWLEDGEMENTS
The authors would like to thank Nazmi El Saleous from the GIMMS group for his help in putting
together this manual, Lorraine Remer from Code 923 for having kindly read the first draft of this
manual and help us to improve it. They are very grateful to Dr. Tucker and Justice from
GSFC/UMD for the logistic support essential to the realization of 6S. Finally, special thanks to all
the Beta tester of the code and in particular to Phil Teillet from Canadian Centre for Remote
Sensing for his interesting and useful remarks. The Laboratoire d'Optique Atmosphrique is
sponsored by CNRS (Centre National de Recherche Scientifique) and by CNES (Centre National
d'Etudes Spatiales).
34
35
APPENDIX I: DESCRIPTION OF THE COMPUTER CODE
This code predicts the satellite signal between 0.25 and 4.0 m assuming cloudless
atmosphere. Let us recall that the apparent reflectance at the satellite level dor Lambertian surface
can be written,
'(
s
,
v
,
v
) t
g
(
s
,
v
){
a
(
s
,
v
,
v
)
T(
s
)
1 (M) S
[
c
(M)e
/
v
(M) t
d
(
v
)]}
This formalism takes into account the main atmospheric effects, gaseous absorption by H
2
O,
O
2
, CO
2
and O
3,
scattering by molecules and aerosols and a non-homogeneous ground may be
considered. The general flow chart for 6S program is reported in Figure A-1
The following input parameters are necessary:
geometrical conditions,
atmospheric model for gaseous components,
aerosol model (type and concentration),
spectral condition,
ground reflectance (type and spectral variation).
At each step, you can either select some proposed standard conditions or define your own
conditions. More details are given in Figures A-2 to A-9. Let us note that the satellite navigation is
done by a rough calculation from the nominal characteristics of the satellite orbit. It is obvious that,
without anchor points, the localization is not very accurate.
From these data, we compute the scattering atmospheric functions, at ten wavelengths which
are well distributed over the solar spectrum (Fig. A-11). At any wavelength, we interpolate from
these values by assuming the spectral dependance
f ( ) f (
0
)
0

_
,

and compute the apparent reflectance from the first equation (Fig.A-10)
We have reported on Fig. A-12 the general outputs and the optional ones.
36
Geometrical Conditions
Aerosol Model (type)
P(,), g(),
o
()
Aerosol Concentration
A
(550 nm)
Spectral Band
inf
,
sup
, S()
Atmospheric Conditions
P(z), T(z), U
H2O
(z), U
O3
(z)
Ground Reflectance
Computation of the Atmospheric functions
at discrete wavelength values
Integration over the spectral band
Computation of the gaseous absorption functions
Solar Spectral Irradiance
Interpolation of the scattering atmospheric functions
Environment reflectance
Computation of the apparent reflectance
Outputs
Computation of BRDF Atmosphere
coupling at equivalent wavelength
Figure A-1 : General flow chart 6S computations
Atmospheric correction if requested
37
I
G
E
O
M
s
,

s
,

v
,

v
,

M
o
n
t
h
,

D
a
y P
O
S
M
T
O
P
O
S
G
E
P
O
S
G
W
P
O
S
N
O
A
P
O
S
S
P
O
P
O
S
L
A
N
N
o
n
e
1
2
3
0
6
7
4
5
s
,

s
,

v
,

v
S
a
t
e
l
l
i
t
e
D
a
t
a
S
u
b
r
o
u
t
i
n
e
F
i
g
u
r
e

A
-
2

:

D
e
t
a
i
l
e
d

F
l
o
w

C
h
a
r
t

f
o
r

G
e
o
m
e
t
r
i
c
a
l

C
o
n
d
i
t
i
o
n
s
N
o
n
e
M
e
t
e
o
s
a
t
G
o
e
s

E
a
s
t
G
o
e
s

W
e
s
t
N
O
A
A

A
M
N
O
A
A

P
M
S
P
O
T
L
a
n
d
s
a
t
M
o
n
t
h
,

D
a
y
,

H
o
u
r
,

N
b
.

c
o
l
s
,

N
b
.

r
o
w
s
M
o
n
t
h
,

D
a
y
,

H
o
u
r
,

L
o
n
g
,

L
a
t
M
o
n
t
h
,

D
a
y
,

H
o
u
r
,

N
b
.

c
o
l
s
,

L
o
n
g

A
.
N
.
,

H
o
u
r

A
.
N
.
38
I
D
A
T
M
0
1
2
3
4
5
6
7
8
N
o
n
e
S
U
B
S
U
M
S
U
B
W
I
N
N
o
n
e
U
S
6
2
P
(
z
)
,

T
(
z
)
,

U
H
2
O
(
z
)
,

U
O
3
(
z
)
S
u
b
r
o
u
t
i
n
e
F
i
g
u
r
e

A
-
3

:

D
e
t
a
i
l
e
d

F
l
o
w

C
h
a
r
t

f
o
r

A
t
m
o
s
p
h
e
r
i
c

C
o
n
d
i
t
i
o
n
s
U
S
6
2
M
I
D
W
I
N
M
I
D
S
U
M
T
R
O
P
I
C
N
o

a
b
s
o
r
p
t
i
o
n
T
r
o
p
i
c
a
l
M
i
d
l
a
t
i
t
u
d
e

S
u
m
m
e
r
M
i
d
l
a
t
i
t
u
d
e

W
i
n
t
e
r
S
u
b
a
r
c
t
i
c

S
u
m
m
e
r
S
u
b
a
r
c
t
i
c

W
i
n
t
e
r
U
S
6
2
U
S
E
R
U
S
E
R
T
y
p
e
N
o
n
e
N
o
n
e
N
o
n
e
N
o
n
e
N
o
n
e
N
o
n
e
N
o
n
e
U
H
2
O
,
U
O
3
D
a
t
a
P
(
z
)
,
T
(
z
)
,
U
H
2
O
(
z
)
,
U
O
3
(
z
)
39
I
A
E
R
0
1
2
3
4
5
N
o
n
e
K
e
x
t
(
)
,

P
(
)
,

g
(
)
,

0
(
)
F
i
g
u
r
e

A
-
4

:

D
e
t
a
i
l
e
d

F
l
o
w

C
h
a
r
t

f
o
r

A
e
r
o
s
o
l

M
o
d
e
l

(
T
y
p
e
)
S
u
b
r
o
u
t
i
n
e
N
o

A
e
r
o
s
o
l
C
o
n
t
i
n
e
n
t
a
l
M
a
r
i
t
i
m
e
U
r
b
a
n
D
e
s
e
r
t
i
c
T
y
p
e
B
i
o
m
a
s
s
S
t
r
a
t
o
s
p
h
e
r
i
c
U
S
E
R

u
s
i
n
g
s
i
z
e

d
i
s
t
r
i
b
u
t
i
o
n
U
S
E
R

u
s
i
n
g
b
a
s
i
c

c
o
m
p
o
n
e
n
t
s
N
o
n
e
N
o
n
e
N
o
n
e
N
o
n
e
N
o
n
e
C
(
i
)

i
=
1
,
4
D
a
t
a
6
7
8

t
o

1
2
N
o
n
e
N
o
n
e
A
E
R
O
S
O
A
E
R
O
S
O
A
E
R
O
S
O
A
E
R
O
S
O
A
E
R
O
S
O
A
E
R
O
S
O
A
E
R
O
S
O
A
E
R
O
S
O
c
f

f
i
g
u
r
e

A
-
4
'
40
S
u
b
r
o
u
t
i
n
e
K
e
x
t
(
)
,

P
(
)
,

g
(
)
,

0
(
)
F
i
g
u
r
e

A
-
4
'

:

D
e
t
a
i
l
e
d

F
l
o
w

C
h
a
r
t

f
o
r

A
e
r
o
s
o
l

M
o
d
e
l

(
T
y
p
e
)
T
y
p
e
D
a
t
a
8
9
1
0
1
1
1
2
M
u
l
t
i

M
o
d
a
l
L
o
g
-
N
o
r
m
a
l
e
d
i
s
t
r
i
b
u
t
i
o
n
(
s
)
(
u
p

t
o

4
)
M
o
d
i
f
i
e
d

G
a
m
m
a
d
i
s
t
r
i
b
u
t
i
o
n
J
u
n
g
e

P
o
w
e
r
-
L
a
w
d
i
s
t
r
i
b
u
t
i
o
n
F
r
o
m

S
u
n
-
p
h
o
t
o
m
e
t
e
r
m
e
a
s
u
r
e
m
e
n
t
s
F
r
o
m

a
n

e
x
t
e
r
n
a
l

f
i
l
e

I
A
E
R
-
U
S
E
R

w
i
t
h

s
i
z
e

d
i
s
t
r
i
b
u
t
i
o
n
r
m
i
n
,
r
m
a
x
,

b
,

r
n
(
)
,

=
1
,
1
0
r
i
(
)
,

=
1
,
1
0
r
m
i
n
,
r
m
a
x
r
n
(
)
,

=
1
,
1
0
r
i
(
)
,

=
1
,
1
0
M
I
E
+
A
E
R
O
S
O
R
e
s
u
l
t
s

s
a
v
e
d

i
n
t
o

t
h
e

f
i
l
e

F
I
L
E
.
m
i
e
I
A
E
R
P
0
1
T
y
p
e
D
a
t
a
F
I
L
E

(
f
i
l
e

n
a
m
e
)
N
o
n
e
i
r
s
u
n
p
h
f
o
r

i
=
1

t
o

i
r
s
u
n
p
h
r
(
i
)
,

(
d
V
/
d
l
o
g
(
r
)
)
(
i
)
d
o
n
e
r
n
(
)
,

=
1
,
1
0
r
i
(
)
,

=
1
,
1
0
r
m
i
n
,

r
m
a
x
,

i
c
p
f
o
r

i
=
1

t
o

i
c
p
r
M
(
i
)
,

(
i
)
,

C
i
j
(
i
)
r
n
(
,
i
)
,

=
1
,
1
0
r
i
(
,
i
)
,

=
1
,
1
0
d
o
n
e
n
a
m
e
o
f

t
h
e

e
x
t
e
r
n
a
l

f
i
l
e
41
V
N
o
n
e
A
(
5
5
0
)
N
o
n
e
-
1
0
A
n
y

p
o
s
i
t
i
v
e

v
a
l
u
e
V
>
5
k
m
N
o
n
e
N
o
n
e
O
D
A
5
5
0
A
(
5
5
0
)
D
a
t
a
S
u
b
r
o
u
t
i
n
e
F
i
g
u
r
e

A
-
5

:

D
e
t
a
i
l
e
d

F
l
o
w

C
h
a
r
t

f
o
r

A
e
r
o
s
o
l

M
o
d
e
l

(
C
o
n
c
e
n
t
r
a
t
i
o
n
,

A
(
5
5
0
)
)
42
X
P
S
P
o
s
i
t
i
v
e

V
a
l
u
e
T
a
r
g
e
t

a
t

s
e
a

l
e
v
e
l
N
e
g
a
t
i
v
e

V
a
l
u
e
T
a
r
g
e
t

a
t

t
h
e
a
l
t
i
t
u
d
e

=

-
x
p
s

(
i
n

k
m
)
P
R
E
S
S
U
R
E
U
H
2
O
,

U
O
3
,
a
e
r
(
5
5
0
)
,
P
(
z
)
,
T
(
z
)
F
i
g
u
r
e

A
-
6

:

D
e
t
a
i
l
e
d

F
l
o
w

C
h
a
r
t

f
o
r

t
h
e

e
f
f
e
c
t

o
f

t
h
e

a
l
t
i
t
u
d
e
:

T
a
r
g
e
t
S
u
b
r
o
u
t
i
n
e
N
o
n
e
T
a
r
g
e
t

A
l
t
i
t
u
d
e
43
-
1
0
0

<

x
p
p

<

0
S
e
n
s
o
r

i
n
s
i
d
e

t
h
e

a
t
m
o
s
p
h
e
r
e
x
p
p

=

-

a
l
t
i
t
u
d
e

(
i
n

k
i
l
o
m
e
t
e
r
s
)
X
P
P
-
1
0
0
0
S
e
n
s
o
r

a
b
o
a
r
d

a

S
a
t
e
l
l
i
t
e
D
a
t
a
S
u
b
r
o
u
t
i
n
e
D
a
t
a
S
e
n
s
o
r

A
l
t
i
t
u
d
e
U
H
2
O
,

U
O
3

b
e
l
o
w

s
e
n
s
o
r
P
R
E
S
P
L
A
N
E
i
f

>

0

o
r

i
f

<

0
a
e
r
(
5
5
0
)

b
e
l
o
w

s
e
n
s
o
r
U
H
2
O
,

U
O
3
,
P
(
z
)
,
T
(
z
)
U
H
2
O
,

U
O
3
,
P
(
z
)
,
T
(
z
)

u
s
i
n
g

U
S

6
2
i
f

>

0

o
r
i
f

<
0
P
R
E
S
P
L
A
N
E

+

U
S

6
2
a
e
r
(
z
)
a
e
r
(
z
)

a
s
s
u
m
i
n
g

s
c
a
l
e

h
e
i
g
h

o
f

2
k
m
0
S
e
n
s
o
r

a
t
G
r
o
u
n
d

l
e
v
e
l
d
e
f
i
n
e
d

b
y

x
p
s
U
H
2
O
,

U
O
3
,
a
e
r
(
5
5
0
)
,
P
(
z
)
,
T
(
z
)
F
i
g
u
r
e

A
-
6
'

:

D
e
t
a
i
l
e
d

F
l
o
w

C
h
a
r
t

f
o
r

t
h
e

e
f
f
e
c
t

o
f

t
h
e

a
l
t
i
t
u
d
e
:

S
e
n
s
o
r
N
o
n
e
N
o
n
e
N
o
n
e
N
o
n
e
N
o
n
e
N
o
n
e
44
I
W
A
V
E
1
2
3

-

4
2
5

-

3
0
i
n
f
,

s
u
p
,

S
(
)
S
u
b
r
o
u
t
i
n
e
F
i
g
u
r
e

A
-
7
:

D
e
t
a
i
l
e
d

F
l
o
w

C
h
a
r
t

f
o
r

S
p
e
c
t
r
a
l

B
a
n
d

D
e
f
i
n
i
t
i
o
n

o
f

E
x
t
r
e
m
e

W
a
v
e
l
e
n
g
t
h
s

a
n
d

F
i
l
t
e
r

F
u
n
c
t
i
o
n
3
1

-

3
4
T
M
H
R
V
1
7

-

2
4
M
S
S
A
V
H
R
R
5

-

1
6
G
O
E
S
M
E
T
E
O
N
o
n
e
N
o
n
e
0
N
o
n
e
-
1
-
2
N
o
n
e
N
o
n
e
N
o
n
e
D
a
t
a
N
o
n
e
N
o
n
e
N
o
n
e
N
o
n
e
o
u
t
p
u
t

p
r
i
n
t
e
d
s
t
e
p

b
y

s
t
e
p

(
2
.
5
n
m
)
i
n
f
,

s
u
p
f
o
r

=
i
n
f
,
s
u
p
,
2
.
5
n
m
S
(
)
d
o
n
e
i
n
f
,

s
u
p
w
i
t
h

S
(
)
=
1
.
0
i
n
f
,

s
u
p
w
i
t
h

S
(
)
=
1
.
0
45
I
H
O
M
O
0
1
n
o
n

h
o
m
o
g
e
n
e
o
u
s
I
D
I
R
E
C
T
I
G
R
O
U
N
I
B
R
D
F
-
1
0
1
2
3
I
G
R
O
U
1
R
A
D
I
G
R
O
U
2
-
1
1
4
3
2
0
R

i
n

k
m
n
o
n
e
n
o
n
e
V
E
G
E
T
A
C
L
E
A
R
W
S
A
N
D
L
A
K
E
W
V
E
G
E
T
A
C
L
E
A
R
W
S
A
N
D
L
A
K
E
W
n
o
n
e
n
o
n
e
0
1
n
o
n
e
R

(
i
n

k
m
)
F
i
g
u
r
e

A
-
8

:

D
e
t
a
i
l
e
d

F
l
o
w

C
h
a
r
t

f
o
r

G
r
o
u
n
d

R
e
f
l
e
c
t
a
n
c
e
,

D
e
f
i
n
i
t
i
o
n

o
f

t
h
e

T
a
r
g
e
t

a
n
d

E
n
v
i
r
o
n
m
e
n
t

R
e
f
l
e
c
t
a
n
c
e
s
S
u
b
r
o
u
t
i
n
e
D
a
t
a
c
f

f
i
g
u
r
e

A
-
8
'
c
(
)
,
e
(
)
( )
,

d
(
)
,

p
(
)
)
4
f
o
r

=
i
n
f
,
s
u
p
,
d
)
w
i
t
h

d
=
2
.
5
n
m
n
o
n
e
n
o
n
e
n
o
n
e
n
o
n
e
h
o
m
o
g
e
n
e
o
u
s
n
o
n
e
n
o
n
e
n
o
n
e
n
o
n
e
c
,
e
c

=

e
f
o
r

=
i
n
f
,
s
u
p
,
d
c
,
e
(
)
w
i
t
h

d
=
2
.
5
n
m
46
I
B
R
D
F
0
1
2
3
4
5
6
7
v
j
,
v
i
)

f
o
r

v
j
,
v
i
)

f
o
r

s
=
v
s
p
h
e
r
i
c
a
l

a
l
b
e
d
o
(
w
i
t
h

j
:
0
0
,

1
0

.
.
.

8
0
,

8
5

i
:
0
0
,

3
0
,

.
.
.

3
6
0
)
(
s
e
e

t
e
x
t

f
o
r

d
e
t
a
i
l
s
)
,
S
(
0
)
,
h
o
p
t
3

o
p
t
4

o
p
t
5
s
t
r
1

s
t
r
2

s
t
r
3

s
t
r
4
o
p
c
.
1

o
p
c
.
2

o
p
c
.
3
(
s
e
e

t
e
x
t

f
o
r

d
e
t
a
i
l
s
)
k
0
,
k
1
,
k
2
a
,
a
'
,
b
,
c
k
,
l
w
s
,
w
,
C
s
a
l
,
C
i
l
d
,
i
h
s
L
t
,
2
r
r
L
,
t
L
,
R
s
i
n
-
s
i
t
u
m
e
a
s
u
r
e
m
e
n
t
s
H
a
p
k
e
'
s
m
o
d
e
l
V
e
r
s
t
r
a
e
t
e
'
s
m
o
d
e
l
R
o
u
j
e
a
n
'
s
m
o
d
e
l
W
a
l
t
h
a
l
l
'
s
m
o
d
e
l
M
i
n
n
a
e
r
t
'
s
m
o
d
e
l
O
c
e
a
n
m
o
d
e
l
I
a
q
u
i
n
t
a

&

P
i
n
t
y
'
s
m
o
d
e
l
v
j
,
v
i
)

f
o
r

v
j
,
v
i
)

f
o
r

s
=
v
s
p
h
e
r
i
c
a
l

a
l
b
e
d
o
(
w
i
t
h

j

=

0
0

t
o

9
0

i

=

0
0

t
o

3
6
0
)
B
R
D
F
G
R
I
D
H
A
P
K
B
R
D
F
H
A
P
K
A
L
B
E
V
E
R
S
B
R
D
F
V
E
R
S
A
L
B
E
R
O
U
J
B
R
D
F
R
O
U
J
A
L
B
E
W
A
L
T
B
R
D
F
W
A
L
T
A
L
B
E
M
I
N
N
B
R
D
F
M
I
N
N
A
L
B
E
O
C
E
A
B
R
D
F
O
C
E
A
A
L
B
E
I
A
P
I
B
R
D
F
I
A
P
I
A
L
B
E
D
a
t
a
S
u
b
r
o
u
t
i
n
e
T
y
p
e
F
i
g
u
r
e

A
-
8
'
:

D
e
t
a
i
l
e
d

F
l
o
w

C
h
a
r
t

f
o
r

B
R
D
F
8
R
a
h
m
a
n
'
s
m
o
d
e
l
0
,
,
k
R
A
H
M
B
R
D
F
R
A
H
M
A
L
B
E
47
RAPP
atmospherically corrected reflectance
Figure A-9 : Detailed Chart Flow for Atmospheric correction mode
RAPP <-1
-1 < RAPP < 0
RAPP > 0
No Atmospheric
Correction
---- Atmospheric Correction mode ----
L
TOA
= RAPP
(in W/m
2
/st/m)
TOA
= - RAPP
using atmospheric conditions previously
described by the user
48
D
I
S
C
O
M
O
D
R
A
Y
L
A
T
M
R
E
F
S
C
A
T
R
A
F
i
g
u
r
e

A
-
1
0

:

D
e
t
a
i
l
e
d

F
l
o
w

C
h
a
r
t

f
o
r

D
i
s
c
r
e
t
e

C
o
m
p
u
t
a
t
i
o
n
s

o
f

t
h
e

S
c
a
t
t
e
r
i
n
g

A
t
m
o
s
p
h
e
r
i
c

F
u
n
c
t
i
o
n
s
,

,

T
(
)

a
n
d

S
49
Loop over wavelength
ABSTRA
SOLIRR
INTERP
ENVIRO
OUTPUTS
VARSOL
Figure A-11 : Detailed Chart Flow for Spectral Integration
50
I
N
T
E
G
R
A
T
E
D

V
A
L
U
E
S

O
F
C
O
U
P
L
I
N
G

A
E
R
O
S
O
L
-
W
A
T
E
R

V
A
P
O
R
I
N
T
E
G
R
A
T
E
D

N
O
R
M
A
L
I
Z
E
D

V
A
L
U
E
S

O
F
%

I
r
r
a
d
i
a
n
c
e

a
t

g
r
o
u
n
d

l
e
v
e
l
R
e
f
l
e
c
t
a
n
c
e

a
t

s
a
t
e
l
l
i
t
e

l
e
v
e
l
F
i
g
u
r
e

A
-
1
2
:

D
e
s
c
r
i
p
t
i
o
n

o
f

t
h
e

o
u
t
p
u
t
s

(
1
/
3
)
%

D
i
r
e
c
t

I
r
r
.
%

d
i
f
f
u
s
e

I
r
r
.
%

E
n
v
i
r
o
.

I
r
r
.
S
(
)
E
s
s
d
S
(
)
E
s
d
S
(
)
E
s
t
d
(
s
)
[
1
s
]
T
(
s
)
d
S
(
)
E
s
d
S
(
)
E
s
e
/
s
[
1
s
]
T
(
s
)
d
S
(
)
E
s
d
A
t
m
.

i
n
t
r
i
n
.

r
e
f
.
E
n
v
i
o
n
m
e
n
t

r
e
f
.
T
a
r
g
e
t

r
e
f
l
e
c
t
a
n
c
e
S
(
)
E
s
T
g
(
s
,
v
)
T
(
s
)
e
/
v
1
s
d
S
(
)
E
s
d
S
(
)
E
s
T
g
(
s
,
v
)
T
(
s
)
t
d
(
v
)
1
s
d
S
(
)
E
s
d
S
(
)
E
s
T
g
(
s
,
v
)
a
d
S
(
)
E
s
d
w
v

a
b
o
v
e

A
e
r
.
:
w
v

m
i
x
e
d

w
i
t
h

A
e
r
.
:
w
v

u
n
d
e
r

A
e
r
.
:
S
(
)
E
s
(
1
)
d
S
(
)
E
s
d
S
(
)
E
s
(
2
)
d
S
(
)
E
s
d
S
(
)
E
s
'
d
S
(
)
E
s
d
A
p
p
a
r
e
n
t

R
e
f
l
e
c
t
a
n
c
e
:
A
p
p
a
r
e
n
t

R
a
d
i
a
n
c
e
:
T
o
t
a
l

g
a
s
e
o
u
s

T
r
a
n
s
m
i
t
t
a
n
c
e
:
S
(
)
E
s
'
d
S
(
)
E
s
d
1
S
(
)
E
s
'
d
1
S
(
)
E
s
T
g
d
(
1
)
T
g
T
g
U
H
2
O
R
+
A
+
(
R
+
A
R
)
T
g
U
H
2
O
+
s *

_

,

w
i
t
h
'
T
g
T
g
U
H
2
O
R
+
A
+
(
R
+
A
R
)
T
g
1 2
U
H
2
O
+
s *

_

,

(
2
)
T
g
T
g
U
H
2
O
R
+
A
+
s *
(
)
w
h
e
r
e
s *
(
s
)
T
(
v
)
s
1
s
s
T
g
U
H
2
O
51
A
B
S
O
L
U
T
E

V
A
L
U
E
S

O
F
I
r
r
a
d
i
a
n
c
e

a
t

g
r
o
u
n
d

l
e
v
e
l

(
W
/
m
2
)
R
a
d
i
a
n
c
e

a
t

s
a
t
e
l
l
i
t
e

l
e
v
e
l

(
W
/
m
2
/
s
t
/
m
)
F
i
g
u
r
e

A
-
1
2
:

D
e
s
c
r
i
p
t
i
o
n

o
f

t
h
e

o
u
t
p
u
t
s

(
2
/
3
)
A
t
m
.

i
n
t
r
i
n
.

r
a
d
.
E
n
v
i
o
n
m
e
n
t

r
a
d
i
a
n
c
e
T
a
r
g
e
t

r
a
d
i
a
n
c
e
1
S
(
)
E
s
T
g
(
s
,
v
)
a
d
1
S
(
)
E
s
T
g
(
s
,
v
)
T
(
s
)
t
d
(
v
)
1
s
d
1
S
(
)
E
s
T
g
(
s
,
v
)
T
(
s
)
e
/
v
1
s
d
D
i
r
e
c
t

S
o
l
a
r

I
r
r
.
A
t
m
.

d
i
f
f
u
s
e

I
r
r
.
E
n
v
i
r
o
.

I
r
r
a
d
i
a
n
c
e
S
(
)
E
s
T
g
(
s
)
T
(
s
)
s
1
s
d
S
(
)
E
s
T
g
(
s
)
t
d
(
s
)
d
S
(
)
E
s
T
g
(
s
)
e
/
s
d
I
n
t
.

F
u
n
c
t
i
o
n

f
i
l
t
e
r

(
i
n

m
)
I
n
t
.

S
o
l
.

S
p
e
c
t
.

(
i
n

W
/
m
2
)
S
(
)
d
S
(
)
s
E
d

52

D
O
W
N
W
A
R
D
U
P
W
A
R
D
T
O
T
A
L
G
a
s

T
r
a
n
s
:
H
2
O
,
C
O
2
,
O
3
,
O
2
N
2
O
,
C
H
4
,
C
O
R
a
y
.

T
r
a
n
s
.
A
e
r
.

T
r
a
n
s
.
T
o
t
.

T
r
a
n
s
.
R
a
y
l
e
i
g
h
A
e
r
o
s
o
l
s
T
o
t
a
l
S
p
h
e
r
i
c
a
l

A
l
b
e
d
o
S
A
S
T
O
p
t
i
c
a
l

D
e
p
t
h
A
R
e
f
l
e
c
t
a
n
c
e
P
h
a
s
e

F
u
n
c
t
i
o
n
P
A
S
i
n
g
l
e

S
c
a
t
t
e
r
i
n
g
A
l
b
e
d
o
1
.
0
F
i
g
u
r
e

A
-
1
0
:

D
e
s
c
r
i
p
t
i
o
n

o
f

t
h
e

o
u
t
p
u
t
s

(
3
/
3
)
S
(
)
s
E
T
g
i
i
1
7
(
s
)
d
S
(
)
s
E
d
S
(
)
s
E
T
g
i
i
1
7
(
v
)
d
S
(
)
s
E
d
S
(
)
s
E
T
g
i
i
1
7
(
s
,
v
)
d
S
(
)
s
E
d
S
(
)
s
E
T
(
s
)
d
S
(
)
s
E
d
S
(
)
s
E
T
(
v
)
d
S
(
)
s
E
d
S
(
)
s
E
T
(
s
)
T
(
v
)
d
S
(
)
s
E
d
S
(
)
S
E
S
R
d
S
(
)
S
E
d
S
(
)
S
E
R
d
S
(
)
S
E
d
S
(
)
S
E
R
d
S
(
)
S
E
d
S
(
)
S
E
P
R
d
S
(
)
S
E
d
R
P
R
+
A
P
A
R
+
A
0
A
A
+
R R
R
+
0
A
A
R
+
A
53
APPENDIX II
EXAMPLE OF INPUTS AND OUTPUTS
INPUT FILE
0 (USER'S CONDITIONS)
40.0 100.0 45.0 50.0 7 23
8 (USER'S MODEL)
3.0 0.35 (UH2O(G/CM2) ,UO3(CM-ATM))
4 (AEROSOLS MODEL)
0.25 0.25 0.25 0.25 (% OF:DUST-LIKE,WATER-SOL,OCEANIC,SOOT)
0 (NEXT VALUE IS THE AER. OPT. THICK. @550)
0.50 (AERO. OPT. THICK. @550)
-0.2 (TARGET AT 0.2 km)
-3.5 (AIRCRAFT AT 3.5 KM)
-1.5 -0.35 (UH2o and UO3 under the plane not avail.)
0.25 (AERO. OPT. THICK. under the plane @550)
11 (AVHRR 1 (NOAA 9) BAND)
1 (GROUND TYPE,I.E. NON UNIFORME SURFACE)
2 1 0.50 (TARGET,ENV.,RADIUS(KM))
-0.10 (ATMOSPHERIC CORRECTION OF RAPP=0.10)
54
OUTPUT FILE (1/3)
******************************* 6s version 4.0 *******************************
* geometrical conditions identity *
* ------------------------------- *
* user defined conditions *
* *
* month: 7 day : 23 *
* solar zenith angle: 40.00 deg solar azimuthal angle: 100.00 deg *
* view zenith angle: 45.00 deg view azimuthal angle: 50.00 deg *
* scattering angle: 146.49 deg azimuthal angle difference: 50.00 deg *
* *
* atmospheric model description *
* ----------------------------- *
* atmospheric model identity : *
* user defined water content : uh2o= 3.000 g/cm2 *
* user defined ozone content : uo3 = .350 cm-atm *
* aerosols type identity : user defined aerosols model *
* .250 % of dust-like *
* .250 % of water-soluble *
* .250 % of oceanic *
* .250 % of soot *
* optical condition identity : *
* visibility : 8.49 km opt. thick. 550nm : .5000 *
* *
* spectral condition *
* ------------------ *
* avhrr 1 (noaa9) value of filter function *
* wl inf= .530 mic wl sup= .810 mic *
* *
* target type *
* ----------- *
* inhomogeneous ground , radius of target .500 km *
* target reflectance : *
* spectral clear water reflectance .045 *
* environmental reflectance : *
* spectral vegetation ground reflectance .129 *
* *
* target elevation description *
* ---------------------------- *
* ground pressure [mb] 989.01 *
* ground altitude [km] .200 *
* gaseous content at target level: *
* uh2o= 3.000 g/cm2 uo3= .350 cm-atm *
* *
* plane simulation description *
* ---------------------------- *
* plane pressure [mb] 657.54 *
* plane altitude [km] 3.500 *
* atmosphere under plane description: *
* ozone content .008 *
* h2o content 1.043 *
* aerosol opt. thick. 550nm .292 *
* *
* atmospheric correction activated *
* -------------------------------- *
* input apparent reflectance : .100 *
* *
*******************************************************************************
55
OUTPUT FILE (2/3)
*******************************************************************************
* *
* integrated values of : *
* -------------------- *
* apparent reflectance .0457 appar. rad.(w/m2/sr/mic) 17.623 *
* total gaseous transmittance .937 *
* *
*******************************************************************************
* *
* coupling aerosol -wv : *
* -------------------- *
* wv above aerosol : .046 wv mixed with aerosol : .046 *
* wv under aerosol : .046 *
* *
* int. normalized values of : *
* --------------------------- *
* % of irradiance at ground level *
* % of direct irr. % of diffuse irr. % of enviro. irr *
* .791 .204 .006 *
* reflectance at satellite level *
* atm. intrin. ref. environment ref. target reflectance *
* .021 .015 .014 *
* *
* int. absolute values of *
* ----------------------- *
* irr. at ground level (w/m2/mic) *
* direct solar irr. atm. diffuse irr. environment irr *
* 617.564 158.453 4.381 *
* rad at satel. level (w/m2/sr/mic) *
* atm. intrin. rad. environment rad. target radiance *
* 7.935 5.846 5.565 *
* *
* *
* int. funct filter (in mic) int. sol. spect (in w/m2) *
* .1174545 185.589 *
* *
*******************************************************************************
56
OUTPUT FILE (3/3)
*******************************************************************************
* *
* integrated values of : *
* -------------------- *
* *
* downward upward total *
* global gas. trans. : .94428 .98766 .93740 *
* water " " : .98570 .99322 .98115 *
* ozone " " : .96462 .99909 .96375 *
* co2 " " : 1.00000 1.00000 1.00000 *
* oxyg " " : .99343 .99533 .99178 *
* no2 " " : 1.00000 1.00000 1.00000 *
* ch4 " " : 1.00000 1.00000 1.00000 *
* co " " : 1.00000 1.00000 1.00000 *
* *
* *
* rayl. sca. trans. : .96434 .98767 .95244 *
* aeros. sca. " : .70714 .80488 .56916 *
* total sca. " : .67887 .79455 .53940 *
* *
* *
* *
* rayleigh aerosols total *
* *
* spherical albedo : .04966 .04532 .06717 *
* optical depth total: .05550 .41782 .47332 *
* optical depth plane: .01848 .24364 .26213 *
* reflectance : .01081 .01323 .02171 *
* phase function : 1.26026 .30427 .41637 *
* sing. scat. albedo : 1.00000 .48167 .54245 *
* *
* *
*******************************************************************************
*******************************************************************************
* atmospheric correction result *
* ----------------------------- *
* input apparent reflectance : .100 *
* measured radiance [w/m2/sr/mic] : 38.529 *
* atmospherically corrected reflectance : .156 *
* coefficients xa xb xc : .00513 .04025 .06717 *
* y=xa*(measured radiance)-xb ; acr=y/(1.+xc*y) *
*******************************************************************************
57
APPENDIX III
DESCRIPTION OF THE SUBROUTINES
Contents
Description of the differents subroutines used to compute the geometrical conditions:
DAY_NUMBER 76
POS_FFT 76
POSGE 63
POSGW 64
POSLAN 65
POSMTO 66
POSNOA 72
POSSOL 76
POSSPO 80
Description of the subroutines used to compute the different atmospheric functions:
ABSTRA 83
AEROSO 91
ATMREF 97
CHAND 99
CSALBR 100
DISCOM 101
DISCRE 102
ENVIRO 103
EXSCPHASE 112
GAUSS 108
INTERP 109
ISO 110
KERNEL 111
58
MIE 112
ODA550 131
ODRAYL 133
OS 135
SCATRA 138
TRUNCA 140
Description of the subroutines used for BRDF ground
BRDFGRID 145
FRESNEL 155
GAULEG 148
GLITALBE 144
HAPKALBE 143
HAPKBRDF 146
IAPIALBE 144
IAPIBRDF 148
IAPITOOLS 148
INDWAT 155
LAD 148
MINNALBE 144
MINNBRDF 152
MORCASIWAT 156
OCEAALBE 144
OCEABRDF 153
OCEATOOLS 153
RAHMALBE 144
RAHMBRDF 163
ROUJALBE 144
ROUJBRDF 165
SOLVE 148
SUNGLINT 154
VERSBRDF 167
VERSALBE 144
VERSTOOLS 167
WALTALBE 144
59
WALTBRDF 169
Description of the subroutines used to update atmospheric profile (airplane, elevated target):
PRESPLANE 175
PRESSURE 176
Description of the different subroutines used to read the data:
AVHRR 181
BBM 201
BDM 201
CLEARW 190
DICA 1 to 3 194
DUST 201
GOES 184
HRV 185
LAKEW 191
METEO 187
METH 1 to 6 195
MIDSUM 202
MIDWIN 203
MOCA 1 to 6 196
MSS 188
NIOX 1 to 6 197
OCEA 201
OXYX 3 to 6 198
OZON 1 199
SAND 192
SOLIRR 179
SOOT 201
STM 201
SUBSUM 204
SUBWIN 205
60
TM 189
TROPIC 206
US 62 207
VARSOL 180
VEGETA 193
WATE 201
WAVA 1 to 6 200
Miscellaneous
EQUIVWL 211
PRINT_ERROR 212
SPECINTERP 213
SPLIE2, SPLIN2, SPLINE, SPLINT 214
61
DESCRIPTION OF THE SUBROUTINES USED
TO COMPUTE THE GEOMETRICAL CONDITIONS
62
63
SUBROUTINE POSGE
Function: Same as POSMTO but for GOES East satellite. We use exactly the same scheme,
only we add the longitude of the subsatellite point, namely 75W, at the retrieval longitude. Let us
also recall that the dimension of the frame is 17331 12997 and the altitude of the satellite is 35729
km.
64
SUBROUTINE POSGW
Function: Same as POSMTO but for GOES West satellite. We use exactly the same scheme
but we add the longitude of the subsatellite point, namely 135W, at the retrieval longitude. Let us
recall that the dimension of the frame is 17331 12997 and the altitude of the satellite is 35769 km.
65
SUBROUTINE POSLAN
Function: To compute the geometrical conditions for the LANDSAT satellite. As the
dimensions of the frame are 180 180 km, the maximum observation angle is 5.5, so we put
V
=
0. The incident conditions are taken from the latitude and the longitude of the centre of the scene.
Reference:
NASA, 1981, GSFC specification for the Thematic Mapper Subsystem and associated test
equipment. Revision C., GSFC 400-8-D.210C, NASA/GSFC, Greenbelt, Maryland, U.S.A..
66
SUBROUTINE POSMTO
Function: To compute the geometrical conditions from the knowledge of the line number and
the pixel in the line (in Meteosat Frame 2500*5000). Firstly, we compute the latitude and the
longitude of the pixel, so the solar position (with the time conditions) can be computed and
secondly, the observation angle.
y
NL
NC
M
W
N
S
P
5000
2500
x
S
x
y
Description: Let S be the satellite, M the subsatellite point, P the observed point, and the
orientation of the axes according to Fig. 1.
From N
c
and N
1
, we obtain the two angles X, Y with respect to x and y axes.
If we refer now to the plane containing the points S, P, M and O the center of the earth (cf.
Fig. 2), we put H the altitude of the satellite and R
E
the earth radius in Equatorial plane.
The observed point P corresponds to the point P' on the earth surface and we have to determine
its coordinates x, y, z with respect to the axis-system centered at O.
67
Figure 2
z
O
H
S
M
N
P
P'
R
E

To obtain z, we put,
z = ON = (R
E
+H) - SN,
SN is obtained from the solutions of triangles OP'S and SNP'.
By solving OP'S, we have
SP'
2
- 2SP' cos OS = OP'
2
- OS
2
,
so
SP' = cos()(H+R
E
)- Error!
By solving SNP' we have SN = SP' cos ,
then SN=cos()
2
(H+R
E
)-cos() (R
E
+H)
2
(cos()
2
-1)-R
E
2

Therefore :
z R
E
+ H cos ( )
2
H+ R
E
R
E
R
E
+ H
R
E

_
,

2
R
E
+ H
R
E

_
,

2
1
cos()
2

1
]
1
1
1
1
To estimate cos, we use the deviations X and Y. A simple trigonometrical identity shows
that :
cos
2
=cos
2
X.cos
2
Y
where
68
cos
2
X=
1
1+tan
2
X

and
cos
2
Y=
1
1+(tanY(1+))
2

with
=
R
E
-R
P
R
E
=
1
297
,
R
E
and R
P
being respectively the equatorial and polar radius, slightly different because of the
earth's oblateness.
O
y
x
z
x1
P"
y
x
z
N
y
To obtain x and y, we consider the Figure 3.
x = -SN tanY
and
y = SN/cosX.tanY
So we have the three coordinates x, y, z of the point P' and infer latitude and longitude .
To compute the longitude, we use, = arctan (x/z) (+ for East, - for West).
To compute the latitude, we have to consider the geoide with semiminor axis R
p
and
semimajor axis R
E
(Figure 4) by solving triangle P'OP".As = atan(y/x
1
)where y is the ordinate of
the point located on the ellipsod, so we have to compute x
1
.
69
Figure 4
R
p
x
1 R
E
x
y
The ellipse equation is written:
x
2
1
R
2
E
+
y
2
R
2
P
=1
so
= atan (
y
R
2
P
-y
2
(R
P
/R
E
)
),
= atan ( tan (R
P
/R
E
))
with = asin(y/R
P
).
To obtain observations angles (azimuthal and zenithal), we use the following simple
geometrical considerations.
For the zenithal angle
v,
70
Figure 5
R
v
P'
S
H
v
= asin ( (1+
h
R
) sin)
where is so as cos
2
= cos
2
X cos
2
Y.
For the azimuthal angle
V
, we solve the spherical triangle P'P''M
Figure 6
v
y
x
z
S

71
where A =
V
- ,
with tan A = tan (1/sin ),
S
v
so
V
= arctan[ tan (1/sin)] + .
From the line and column numbers in a MTOSAT Frame, we can compute the latitude ()
and the longitude () of the point, and the viewing direction from the normal at the point (azimuthal
V
and zenithal
V
angles). Moreover, if we know the date and the hour of the acquisition, we can
obtain the solar conditions (
s,

s
) from the subroutine POSSOL.
Reference:
MORGAN, 1981, Introduction to the Meteosat System, ESOC, Darmstadt, Germany.
72
SUBROUTINE POSNOA
Function: To compute the geometrical conditions for the NOAA series satellites. Generally,
we know the pixel number on a line, the longitude and the time of the ascendant node at the
equator, and the time of the acquisition. We obtain latitude and longitude of the viewed point, the
viewing angles and with the knowledge of the date, the solar geometrical conditions.
Description: The altitude of NOAA satellite is about H = 860 km, the orbit inclination
98.96and the time of one revolution is about 101.98 mn (6119 sec.). The 1/2 angle is of maximum
55.385and you have 2048 pixels for each line.
Figure 1
Gre. Mer.
P
Eq.
S
N
Na
98.96
Let AN be the hour movement in rad/sec, H
N
the hour at the ascendant node,
N
its longitude
and N
c
the pixel number.
73
Figure 2
S
N
Na
N
R
Na
i
P
U
B
S
G.M.
v
Consider Figs.2, S is the subsatellite point, N the ascendant node and P the observed point.
The scan angle gives an angle noted at the centre of the earth.
By solving the triangle PRN, we have the latitude
P
so as:
sin
P
= sin (i+B) sin (NP)
Now, in triangle PSN,
sin (NP) = sin / sin B
and
tan (B) = tan / sin U
So,
sin
P
= cos i sin +sin i cos sinU
By solving PRN, we obtain the longitude
P
with respect to
N,
sin
P
= sin NPR sin NP
74
with
sin NPR =
cos(i+B)
cos
P
,
so, we write
sin
P
=
cos(i+B)sinNP
cos
P

or
sin
P
=
cos(i+B)sin
cos
P
sinB

sin
P
=
-sinisin+coscosisinU
cos
P
.
To completly determine the longitude, we use the other relation which gives the cosine
cos
P
=
coscosU
cos
P

The absolute longitude (Greenwich, Meridian reference) is given by
=
P
+
N -
(T-H
Na
)

2
86400

,
where T is the time of the acquisition, the last term is for taking the rotation of the earth between T
and H
Na
into account. Let us recall that the mouvement angle U is calculated from U=AN.(T- H
Na
).
Consider again Figs. 2 to determine the azimuthal and zenithal observation angles.
is so as
= 55.385
N
c
-1024
1024
in deg.
the zenithal viewing angle
V
is defined by
V
=asin[(1+
H
R
)sin]
The observation azimuthal angle
V
is determinated by solving the triangle NSP,
sin
V
=
sin(
S
-
P
)cos
s
sin

75
and
cos
V
=
sin
S
-sin
P
cos
cos
P
sin

where
S
and
S
are the latitude and the longitude of the subsatellite point P.
Reference:
The characteristics of the orbit have been taken from, NOAA POLAR ORBITER DATA USERS
GUIDE, 1985, U.S. Department of Commerce, NOAA: National Environment Satellite,
National Climatic Data Center, Satellite Data Service Division, World Weather Building,
Room 100, Washington DC 20233, U.S.A..
76
SUBROUTINE POSSOL
Function: To compute the solar azimuthal and zenithal angles (in degrees) for a point over
the globe defined by its longitude and its latitude (in dec. degrees) for a day of the year (fixed by
number of the month and number of the day in the month) at any Greenwich Meridian Time (GMT
dec. hour).
Figure 1
S
N
t
S
P
Description: Let P be the point determined by the latitude and the declination of the sun
at this period of the year, the hour angle is noted t. So the incident angle
S
can be determined by
spherical trigonometry expression
77
cos
S
= cos (
2
- ) cos (
2
- ) + sin (
2
- ) sin (
2
- ) cos t
or
cos
v
= sin sin + cos cos cos t
The solar declination depends upon the day of the year. We used the decomposition in Fourier
series of the declination based on astronomical data with the expression:
=
1
-
2
cos(A)+
3
sin(A) -
4
cos(2A)+
5
sin(2A)-
6
cos(3A)+
7
sin(3A)
where A=
2J
365
and J is the julian day
1
=.006918,
2
=.399912,
3
=.070257,
4
=.006758,
5
=.000907,
6
=.002697,
7
=.001480
The hour angle is computed from the following considerations. From the GMT time, we
compute the mean solar time (or local time) for the longitude
MST = GMT +
15
(dec.hour) .
The length of the day changes within the year (differences between +30 s and - 20 s), so we
have to correct the local time to obtain the true solar time (TST).
TST = MST + ET
where the equation of time ET is given by:
ET=
(
1
+
2
cos(B)-
3
sin(B)-
4
cos(2B)-
5
sin(2B))12
(dec.hour)
with
B =
2J
365
,
1
=.000075,
2
=.001868,
3
=.032077,
4
=.014615,
5
=.040849
We obtain the hour angle t
t = 15
180
(TST-12) (radians)
and can compute
S
.
78
Figure 2
N
t
S P
s
/2 - /2 -

To determine the azimuthal angle
S
, we solve the spherical triangle NSP:
sin
sin(
2
-)
=
sint
sin
S

where is the solar azimuthal angle measured from the south through the west.
Or
sin=cos
sint
sin
S
.
To determine the sign of we use the cosine
cos=
cossin+coscost
sin
S

so is completly defined.
To define the solar azimuthal angle
S
with respect to North, we write,
S
= + asin .
References:
Ch. PERRIN DE BRICHAMBAUT, Rayonnement Solaire et Echanges Radiatifs Naturels,
Monographies de Mtorologie, Gauthier-Villars, Paris, France, 1963.
N. ROBINSON, Solar Radiation, Elsevier Publishing Company, New-York, N.Y., 10017, 1966.
Figure 3: Simulation of solar angles for the 1
rst
May, at different Latitudes, versus Universal Time.
79
0
20
40
60
80
100
120
140
160
180
0 4 8 12 16 20 24
Lat: 45 south
Long: 0
S
o
l
a
r

z
e
n
i
t
h

a
n
g
l
e
Universal time
0
40
80
120
160
200
240
280
320
360
0 4 8 12 16 20 24
Long: 0
S
o
l
a
r

a
z
i
m
u
t
h

a
n
g
l
e
Universal time
Lat: 45 south
0
20
40
60
80
100
120
140
160
180
0 4 8 12 16 20 24
Long: 0
S
o
l
a
r

z
e
n
i
t
h

a
n
g
l
e
Universal time
Lat: 10 south
0
40
80
120
160
200
240
280
320
360
0 4 8 12 16 20 24
Long: 0
S
o
l
a
r

a
z
i
m
u
t
h

a
n
g
l
e
Universal time
Lat: 10 south
0
20
40
60
80
100
120
140
160
180
0 4 8 12 16 20 24
Long: 0
S
o
l
a
r

z
e
n
i
t
h

a
n
g
l
e
Universal time
Lat: 45 north
0
40
80
120
160
200
240
280
320
360
0 4 8 12 16 20 24
Long: 0
S
o
l
a
r

a
z
i
m
u
t
h

a
n
g
l
e
Universal time
Lat: 45 north
80
SUBROUTINE POSSPO
Function: To compute the geometrical conditions for the SPOT satellite. As the dimensions
of the frame are 60 60 km with an observation angle of maximum 2.06, we have considered that:
the zenithal observation angle is nul, so the azimuthal angle is not defined,
the incident conditions are the same that those computed for the center of the frame.
Note: We have not considered the off nadir viewing.
Reference:
M. CHEVREL , M. COURTOIS, G. WEILL (1981). The SPOT Satellite Remote Sensing Mission,
Photogrammetric Engineering and Remote Sensing, 47, 1163-1171.
81
DESCRIPTION OF THE SUBROUTINES USED
TO COMPUTE THE ATMOSPHERIC CONDITIONS
82
83
SUBROUTINE ABSTRA
Function: To compute the gaseous transmittance between 0.25 and 4 m for downward,
upward and total paths. We consider the six gases (O
2
, CO
2
, H
2
O, O
3
, N
2
O and CH
4
) separately.
The total transmission is put equal to the simple product of each ones. The spectral resolution is
equal to 10 cm
-1
.
Description: We have used two random exponential band models (Goody for H
2
O and
Malkmus for O
2
, CO
2
, O
3
, N
2
O and CH
4
) to compute the gaseous transmissions. If we consider an
homogeneous path, the transmission function is written,
for H
2
O
t
G
exp
N
0
k m
1 +
k m

0

_
,

1
2

1
]
1 (01)
for the other gases
t
M
exp
2
0
N
0
1 +
k m

0

_
,

1
2
1

_
,

1
]
1
1
, (02)
where m is the absorber amount, N
0
the total line number in the frequency interval , k the
average intensity and
0
the average Lorentz half width, obtained from intensity Sj and half width
j of the j
th
spectral line by
k
S
j
j 1
N
0
N
0
(03)
k
1
4
S
j
j 1
N
0
S
j
j ( )
1/2
j1
N
0

1
]
1
1
1
2
(04)
84
The spectral resolution of 10 cm
-1
is sufficient and contains enough spectral lines to use a
random band model transmission function.
From a general point of view, the width of a spectral line corresponds to the convolution
product of the two shapes, Lorentz and Doppler and is therefore called a Voigt line. For an
atmospheric gas (O
2
, CO
2
, H
2
O, O
3
, N
2
O and CH
4
) the altitude where the Lorentz width and
Doppler width are equivalent, is about 30 km. So, according to the vertical distribution, only O
3
requires a more complex treatment to take into account a Voigt profile. The O
3
visible transmission
is computed by an other method detailed in the next part and the absorption in the solar infrared (3.3
m) is very small (cf. Fig. I-2 of the chapter 1, 1). Therefore, we have used the same formalism for
all gases. The approximation contributes no consequential error.
Equations (1) and (2) are valid for a homogeneous path, where pressure and temperature are
assumed to be constant. To take into account the variations of temperature and pressure along the
atmospheric path, we use the Curtis-Godson approximation which associates an amount m
weighted by temperature (thereby related to the line intensity), and a amount m weighted by
pressure and temperature (thereby related to the intensity and half width line)
m (z,z') (T) du
z
z'
, (05)
m(z,z') (T) du
z
z'
, (06)
with
=p/p
0
(p
0
is the standard pressure at which the measurements of spectroscopic parameters have
been made)
du=g(dz/) (g is the gaseous density and the cosine of the viewing angle).
The functions (T) and (T) are given by
(T)
S
j
(T)
j1
N
0
S
j
(T
r
)
j 1
N
0
, (07)
85
(T)
S
j
(T)
j0
(T)
[ ]
1/2
j 1
N
0
S
j
(T
r
)
j0
(T
r
)
[ ]
1/2
j1
N
0

1
]
1
1
1
2
,
(08)
with T
r
the reference temperature and j
0
the half-width at temperature T
r
and pressure p
0
.
To simplify, we fit these functions with,
(T) exp a(T T
r
) + b(T T
r
)
2
[ ]
, (09a)
(T) exp a'(T T
r
) + b'(T T
r
)
2
[ ]
. (09b)
The spectroscopic data are taken from the AFGL atmospheric absorption line parameters
compilation (1991 edition). We have selected the following parameters,
the position (in cm
-1
),
the integrated line strength Sj (T
r
) at 296 K (in cm
-1
/(molecules-cm
2
)),
the half width j
0
at 296 K and 1013 mb (in cm
-1
),
the energy of the lower transition state.
The half width at any temperature and pressure is obtained by
j
(p,T)
j0
p
p
0

_
,

T
r
T

_
,
1/2
(10)
and the intensity at any temperature can be computed from the vibrational and rotational partition
and the energy of the lower transition state.
Subsequently, we have taken T
r
= 250 K and computed (T) and (T) for 3 temperatures
(200, 250 and 300 K) to determine the coefficients a, a', b and b'.
Now we have a series of eight coefficients by steps of 10 cm
-1
:
86

k
S
j
j 1
N
0
(T
r
)
(11)

S
j
(T
r
)
j0
(T)
[ ]
1/2
j1
N
0

1
]
1
2
S
j
(T
r
)
j 1
N
0
(12)
a, a'
b, b'

low
the lower frequency of the interval, and
sup
=
low
+ 10 cm
-1
These coefficients are read in the subroutines, WAVA1 to 6 for H
2
O, OZON1 for O
3
,
OXYG3 to 6 for O
2
and DICA1 to 3 for CO
2
The weighted absorber amounts m and m are computed according to Eq. (5) and (6) and the
transmission functions (which correspond to Eq. (1) and (2) for a homogeneous path) are written
t
G
exp
k m
1 +
k

0
m
2
m

_
,

1
2

1
]
1
1
, (13)
t
M
exp

0
m
2m
1 +
4 k

0
m
2
m

_
,

1
2
1

_
,

1
]
1
1
. (14)
Due to the deficiency of spectroscopy data, the visible ozone transmission function is written,
t
O
3
() exp(A
O
3
() u
O
3
) (15)
where u
O
3
is the absorber amount, A
O
3
the absorption coefficient given by Kneizys and al. (1980).
These coefficients are given in steps of 200 cm
-1
between 13000 and 24200 cm
-1
and by step of
500 cm
-1
between 27500 - 50000 cm
-1.
87
To take into account the water vapor continuum, we use the same expression with the
coefficients A
H
2
O
c
are given in step of 5 cm
-1
between 2350 and 2420 cm
-1
.
A comparison between MODTRAN2 and our results (6S) is shown in the following figures
(1-3). The difference observed at roughly 3.1m is due to the fact that we have not taken into
account the N
2
O continuum. This spectral range is already contaminated by water vapor and is not
an atmospheric window. Therefore, the 3.1m region is not used in remote sensing and its emission
in 6S generally unimportant.
References:
A.R. CURTIS, The computation of radiative heating rates in the atmosphere, Proc. Roy. Soc.
London, A236, p. 156-159, 1956.
R.G. ELLINGSON, J.C. GILLE, An infrared radiative transfer model. part 1: Model description
and comparison of observations with calculations, J. Atmos. Sci. 35, p. 523-545, 1978.
R.M. GOODY, Atmospheric Radiation 1, Theoretical Basis, Oxford University Press, 436 pp, 1964.
F.X. KNEIZYS, E.P. SHETTLE, W.O. GALLERY, J.H. CHETWYND Jr.L.W. ABREU, J.E.A.
SELBY, R.W. FENN, R.A. Mc CLATCHEY, Atmospheric Transmittance/Radiance:
Computer Code LOWTRAN 5, AFGL-TR-80-0067, Air Force Geophysics Laboratory,
Bedford, Mass. 1980.
W. MALKMUS, Random Lorentz Band Model with Exponential- Tailed S
-1
Line-intensity
Distribution Function, J. Opt. Soc. Am. 57, 3, p. 323-329, 1967.
J.J. MORCRETTE, Sur la Paramtrisation du Rayonnement dans les Modles de la Circulation
Gnrale Atmosphrique, Thse d'Etat no 630, Universit de Lille.
D.C. ROBERTSON, L.S. BERNSTEIN, R. HAIMES, J. WUNDERLICH, L. VEGA, 5 cm
-1.
Band
Model Option to Lowtran 5, Applied Optics, 20, p. 3218-3226, 1981.
88
C.D. RODGERS, C.O. WALSHAW, The Computation of Infrared Cooling Rates in Planetary
Atmospheres, Quart. J. Roy. Meteor. Soc., 92, p. 67-92, 1966.
L.S. ROTHMAN, R.R. GAMACHE, A. BARBE, A. GOLDMAN, J.R. GILLIS, L.R. BROWN,
R.A. TOTH, J.M. FLAUD, C. CAMY-PEYRET, AFGL Atmospheric Absorption Line
Parameters Compilation: 1982 Edition, Applied. Optics, 22, p. 2247-2256, 1983.
89
0.0
0.2
0.4
0.6
0.8
1.0
0.2 0.4 0.6 0.8 1.0 1.2
6S
MODTRAN
T
g
Wavelength (m)
Figure 1: Gaseous transmission between 0.25 and 1.2m (mid. lat. summer atmosphere)
0.0
0.2
0.4
0.6
0.8
1.0
1.2 1.4 1.6 1.8 2.0 2.2 2.4
6S
MODTRAN
T
g
Wavelength (m)
Figure 2: Gaseous transmission between 1.20m and 2.40m (mid. lat. summer atmosphere).
90
0.0
0.2
0.4
0.6
0.8
1.0
2.4 2.6 2.8 3.0 3.2 3.4 3.6 3.8 4.0
6S
MODTRAN
T
g
Wavelength (m)
Figure 3: Gaseous transmission between 2.40m and 4.00m (mid. lat. summer atmosphere).
91
SUBROUTINE AEROSO
Function: To compute the optical scattering parameters (extinction and scattering
coefficients, single scattering albedo, phase function, assymetry factor) at the ten discrete
wavelengths for the selected model (or created model) from:
(1) the characteristics of the basic components of the International Radiation Commission.
(1983).
dust-like component (D.L., SUBROUTINE DUST)
oceanic component (O.C., SUBROUTINE OCEA)
water-soluble component (W.S., SUBROUTINE WATE)
soot component (S.O., SUBROUTINE SOOT)
(2) pre-computed caracteristics,
now available are the desertic aerosol model corresponding to background conditions, as
described in Shettle(1984), a stratospheric aerosol model as measured Mona Loa (Hawaii)
during El Chichon eruption and as described by King et al. (1984), and a biomass burning
aerosol model as deduced from measurements taken by sunphotometers in Amazonia.
(SUBROUTINES BDM, STM and BBM)
(3) computed using the MIE theory with inputs (size distribution, refractive indexes...) given
by the user (see SUBROUTINES MIE and EXSCPHASE).
These models don't correspond to a mixture of the four basic components.
Description: From the MIE theory (see SUBROUTINE MIE), we have computed the phase
function P(), the extinction and scattering coefficients, the assymetry factor g for the basic
components defined by their size distributions and their refractive index. The computations were
performed at 10 wavelengths and 83 phase angles (80 Gauss angles, 0, 90, 180)
Note: We compute the resultant phase function for the scattering angle by linear interpolation
in the table of 83 values.
From the four basic components, three tropospheric aerosols types models have been selected
by mixing with the following volume percentages. By mixing, we suppose an idea of "external
92
mixing" in the model construction, so the resultant values are obtained by a weighted average using
the volume percentages Cj given by:
D.L. W.S. O.C. S.O.
Continental 0.70 0.29 0.01
Maritime 0.05 0.95
Urban 0.17 0.61 0.22
For each component, we know the volume concentration Vj and the particle number
concentration Nj (particle/cm
3
):
D.L. W.S. O.C. S.O.
Vj
3
/cm
3
113.98352 113.98352 10
06
5.14441 59.777553 10
06
Nj part/cm
3
54.73400 1.86850 10
+06
276.0500010 1.805820 10
+06
where
V
j
4
3
r
3
dN
j
(r)
dr
0
+
dr
and Nj is computed so as to normalize the extinction coefficient at 550 nm .
If Cj is the aerosol fraction by volume of the component j, we have Cj =vj / v with vj = nj Vj
where nj is the number of particles in the mixing so
n n
j
v
C
j
V
j j
then we can obtain the percentage density of particles

n
j
n
C
j
V
j
C
j
V
j j
so for example nj/n for the 3 selected models:

93
D.L. W.S. O.C. S.O.
Continental 2.26490 10
06
0.938299 0.0616987
Maritime 0.999579 4.20823 10
04
Urban 1.65125 10
07
0.592507 0.407492
To obtain the extinction coefficient of the resultant model, we compute
K
ext
()
n
j
n
K
j
ext
()
j
and we normalize also this coefficient at 550 nm. So we have to compute the equivalent
number N of particles by :
N
1
n
j
n
K
j
ext
(550)
j
Since K
ext
j
(550) = 1/Nj, we obtain
1
N
n
j
n
j
1
N
j
The other optical parameters are computed by the same way:
scattering coefficient :
K
sca
() N
n
j
n
j
K
j
sca
()
assymetry factor :
g()
N
K
sca
()
n
j
n
j
g
j
() K
j
sca
()
phase function :
P
()
N
K
sca
()
n
j
n
j
P
j
()K
j
sca
()
the single scattering albedo is directly obtained by the ratio
0
()
K
sca
()
K
ext
()
94
Notes:
- The data for extinction or scattering coefficients are in km
-1
- The following figures give us an order of magnitude of these terms for the 3 selected aerosol
models plus the desert aerosol model.
Reference:
World Meteorological Organization (CAS)/Radiation Commission of IAMAP Meeting of experts
on aerosols and their climatic effects, WCP 55, Williamsburg, Virginia, U.S.A., 28-30 March
1983.
E. P. SHETTLE, Optical and radiative properties of a desert aerosol model. Symposium of
Radiation in the atmosphere, 1 Deepak publishing), pp. 74-77, 1984.
M. KING, HARSHVARDHAN, and ARKING, A, A model of the Radiative Properties of the El
Chichon Stratospheric Aerosol Layer, J. Appl. Meteor., 23, (7), pp. 1121-1137, 1984.
95
Figure 1: Spectral dependence of the extinction coefficient for various aerosol models.
0.01
0.1
1
0.4 0.6 0.8 1.0 2.0 3.0
Continental
Maritime
Urban
Desertic
Biomass
Stratospheric
E
x
t
i
n
c
t
i
o
n

C
o
e
f
f
i
c
i
e
n
t
Wavelength (mm)
Figure 2: Spectral dependence of the single scattering albedo for various aerosol models.
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
0.4 0.6 0.8 1.0 2.0 3.0
Continental
Maritime
Urban
Desertic
Biomass
Stratospheric
S
i
n
g
l
e

S
c
a
t
t
e
r
i
n
g

A
l
b
e
d
o
Wavelength (mm)
96
97
Figure 3: Spectral dependence of the assymetry parameter for various aerosol models.
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.4 0.6 0.8 1.0 2.0 3.0
Continental
Maritime
Urban
Desertic
Biomass
Stratospheric
A
s
s
y
m
e
t
r
y

P
a
r
a
m
e
t
e
r
Wavelength (m)
Figure 4: Phase function at 0.550m versus scattering angle for various aerosol models.
0.01
0.1
1
10
1
10
2
10
3
0.0 30.0 60.0 90.0 120.0 150.0 180.0
Continental
Maritime
Urban
Desertic
Biomass
Stratospheric
P
h
a
s
e

F
u
n
c
t
i
o
n
Scattering Angle ()
98
99
SUBROUTINE ATMREF
Function: To compute the atmospheric reflectance for the molecular and aerosol atmospheres
and the mixed atmosphere. In 6S instead of an approximation as in 5S, we use the scalar Successive
Order of Scattering method (subroutine OS.f). The polarization terms of aerosol or rayleigh phase
are not accounted for in the computation of the aerosol reflectance and the mixed Rayleigh-aerosol
reflectance. The polarization is addressed in computing the Rayleigh reflectance (Subroutine
CHAND.f) by semi-empirical fitting of the vectorized Sucessive Order of Scattering method (Deuz
et al, 1989).
Description: Three reflectance terms have to be computed by ATMREF.f, the aerosol
reflectance (
A
), the rayleigh reflectance (
R
) and the reflectance of the mixed Rayleigh-aerosol
(
R+A
). In addition three different configurations of sensor position are possible, ground based
observation, satellite sensor or airborne sensor.
In the case of ground based observations, we consider that there is no contribution of the
atmosphere below the sensor and the three reflectances are simply set to zero.
For the case of satellite based observations, we can consider that all the molecules and aerosol
are below the sensor. In that case, we use the subroutine OS.f to compute
A
and
R+A
and the
subroutine CHAND.f to compute
R
. The subroutine OS.f is able to deal with a mixture of
molecules and aerosols or with aerosol only or molecules only, by computing the signal in a set of
layers for which the proportion of molecules and aerosol can be adjusted. The computation of the
proportion of aerosol and molecules in each layer is optimized by the subroutine DISCRE.f to
divide the entire atmosphere in equal optical depth layers, the proportion depends on the aerosol
profile which is assumed to be exponential with a scale heigh of 2km.
100
For the case of airborne observation, the three components are computed by the OS.f
subroutine. In OS.f, a special layer is set so that the top of the layer corresponds to the aircraft's
altitude. When aerosol optical depth below the plane is provided by the user as encouraged, the
scale height of aerosol is computed again to match the total aerosol optical depth, the aerosol optical
below the plane and the plane altitude. If in that case, the scale heigh is found to be greater than
7km a warning message is issued and computation are aborted.
References:
Radiation Commission of IAMAP, Standart Procedures to compute Atmospheric Radiative
Transfer in a scattering Atmosphere. Edited by J. LENOBLE, Available from Dr. S. Ruttenberg,
NCAR, Boulder Colorado 80307, U.S.A., 1977.
D. TANRE, M. HERMAN, P.Y. DESCHAMPS, A. DE LEFFE, Atmospheric Modeling for Space
Measurements of Ground Reflectances including bidirectional properties, Appl. Opt. 18, no 21,
p. 3587-3594, 1979.
J. L. DEUZ, M. HERMAN AND R. SANTER, Fourier series expansion of the transfer equation in
the atmosphere-ocean system. Journal of Quantitative Spectroscopy and Radiative Transfer, 41,
6, 483-494, 1989.
101
SUBROUTINE CHAND
Function: To compute the atmospheric reflectance for the molecular atmosphere in case of satellite
observation.
Description: In 6S, to save computer ressources but maintain a good accuracy, we used the same
approach to compute the molecular scattering reflectance as it is detailled in a recent paper
(Vermote and Tanr 1992). The molecular reflectance, as computed from 6S is plotted vs the
reflectance computed from the SOS method for =0.35 in Fig. 1. Four values of the solar zenith
angle (0, 53, 66 and 70), 17 values of the viewing zenith angle (from 0 to 60 with a step of
3.3) and 19 values of the difference of the azimuth angles (from 0 to 180 with a step of 10),
covering a large range of possible geometrical conditions, have been selected. Multiple points fall
on the 45-degree line; the right-hand scale, which gives absolute differences between the two
results, clearly shows that the accuracy of 0.001 is achieved for the full range of geometric
conditions.
References:
E. F. VERMOTE and D. TANRE , Analytical Expressions for Radiative Properties of Planar
Rayleigh Scattering Media Including Polarization Contribution. Journal Of Quantitative
Spectroscopy and Radiative Transfer, 47, 4, 305-314, 1992.
Figure 1: Accuracy of CHAND.f
0.10
0.20
0.30
0.40
0.50
0.60
0.000
0.001
0.002
0.003
0.004
0.005
0.10 0.20 0.30 0.40 0.50 0.60
Ref lect ance
Absolut e Dif f erence
A
P
P
R
O
X
I
M
A
T
E

R
E
F
L
E
C
T
A
N
C
E
ACTUAL REFLECTANCE
A
B
S
O
L
U
T
E

D
I
F
F
E
R
E
N
C
E
102
SUBROUTINE CSALBR
Function: To compute the spherical albedo of the molecular layer.
Description: We integrate the transmission function of the different incident directions to
calculate the spherical albedo, s, that is:
s 1 T() d
0
1
(01)
Using the expression of T() derived in SCATRA (Eq. 01.), it can be shown that s reduces to:
s
1
4 + 3
[3 4E
3
() + 6E
4
()] (02)
where E
3
() and E
4
() are exponential integrals for the argument . These functions are easily
computable from expressions given in the reference below.
Figure 1 shows that the differences between the exact results and Eq. (02) are approximately
0.003 for 0.35 which results in an error of 0.0003 for a surface albedo of 0.10. In the red part of
the solar spectrum for which the surface albedo may be larger, the error is still below 0.001.
References
M. ABRAMOWITZ AND I STEGUN, Handbook of Mathematical Functions (New-York: Dover
Publications,Inc), 1970.
0.00
0.05
0.10
0.15
0.20
0.25
0.000
0.002
0.004
0.006
0.008
0.010
0.05 0.10 0.15 0.20 0.25 0.30 0.35
exact result s
analyt ical expression
Er r or
S
P
H
E
R
I
C
A
L

A
L
B
E
D
O
A
B
S
O
L
U
T
E

D
I
F
F
E
R
E
N
C
E

OPTICAL THICKNESS
Figure 1: Accuracy of Eq. 02.
103
SUBROUTINE DISCOM
Function : To compute the optical properties of the atmosphere at the 10 discrete
wavelengths.
Description :
The 10 wavelengths, 0.400, 0.488, 0.515, 0.550, 0.633, 0.694, 0.860, 1.536, 2.250, 3.750,
have been selected because they correspond to the atmospheric windows used in remote sensing.
The computed quantities are
molecular optical depth (subroutine ODRAYL)
aerosol optical depth(subroutine ODA550)
atmospheric reflectances (subroutine ATMREF)
scattering transmittances (subroutine SCATRA)
spherical albedos (subroutine SCATRA).
The computations have been made respectively for the 3 types of atmosphere:
molecular
aerosols only
complete atmosphere with the two components.
104
SUBROUTINE DISCRE
Function: Decompose the atmosphere in a finite number of layers. For each layer, DISCRE
provides the optical thickness, the proportion of molecules and aerosols assuming an exponential
distribution for each constituants. Figure 1 illustrate the way molecules and aerosols are mixed in a
realistic atmosphere. For molecules, the scale height is 8km. For aerosols it is assumed to be 2km
unless otherwise specified by the user (using aircraft measurements).
Figure 1: Molecules and aerosol mixing in atmosphere.
molecule - aerosol cont ent
l ayer 2
l ayer 3
ground
sur f ace
molecule
aer osol
=0
=1
(z)=(0)exp(-z/h)
105
SUBROUTINE ENVIRO
Function: To compute the environment functions F(r) which allows us to account for an in-
homogeneous ground.
Description: For an accurate evaluation of F(r), Monte Carlo computations are necessary to
take into account
the altitude dependence of the phase function
the dependence of the phase function upon the aerosols type
the scaling factors which are different for the aerosols and the molecules.
Simulations for some different vertical distributions and phase functions show that the
variability of the environment function F(r) can be rather tractable.
The molecular scattering which is a major factor for the enlarged contribution of the
background can be linearized and accounted for by:
F(r)
t
d
R
(
v
)F
R
(r) + t
d
P
(
v
) F
P
(r)
t
d
R
(
v
) + t
d
P
(
v
)
(01)
where t
d
R
(
v
) and t
d
P
(
v
) are the diffuse fractions in the transmission functions respectively
for Rayleigh and aerosols.
F
R
(r) and F
P
(r) correspond to the environment functions estimated for Rayleigh and aerosols
taken into account separately, these functions are slightly dependent upon the wavelength.
We have computed these 2 functions for a mean atmosphere (Mc Clatchey et al, 1971) and we
propose the following approximations:
F
P
(r) 1 0.375 E
0.20r
+ 0.625E
1.83r
[ ]
(02)
F
R
(r) 1 0.930 E
0.08r
+ 0.070 E
1.10r
[ ]
(03)
where r is in km.
106
If the actual aerosol model (type and vertical distribution) does not differ much from the mean
model, these approximations are reasonnable and we account for major part of the environment
effect. Figure 1 shows the two functions F
R
(r) and F
P
r. We note that the horizontal scales of the
environment effect are typically 1km for aerosol scattering and 10km for molecular scattering.
For the case of an airborne observation, we computed the altitude dependence of the
Rayleigh and aerosol environment function. For several typical altitude we have computed F
R
(r,z)
and F
P
(r,z) by the Monte Carlo method and we have derived an approximate expression (Eq. (2)
and (3)). Figures 2 and 3 show, for the selected altitudes, the environment functions For a plane
flying at an arbitrary altitude, we perform a linear interpolation between the closest simulated
altitudes in 6S to get the environment function at the altitude of the plane.
Effect of the view zenith angle.
For 6S, we look at the dependence of these two environment functions as a function of the
view zenith angle. Fig. 4a and 4b show for several values of the view zenith angle the environment
function of Rayleigh and aerosol. As it can be observed on Fig. 4a-b, there is a dependence of the
function F(r) on the view direction for view zenith angle larger than 30. In order to account for
this effect, we chose to fit the environment function at the desired view angle solely as a function of
the environment function computed for a nadir view as it is suggested by Fig. 4a-b. The results
presented on Fig. 4a-b (symbols) show that a simple polynomial function of nadir view
environment function whose coefficients depend on the logarithm of the cosine of view angle is
adequate. For molecules, the F function is fitted by the simple expression:
F
R
(
v
) F
R
(
v
0). ln(cos(
v
). 1 F
R
(
v
0) ( ) +1
[ ]
(04)
for aerosol, a polynomial of a higher degree is needed, that is:
F
A
(
v
) F
A
(
v
0).
1+ a
0
ln(cos()) + b
0
ln(cos())
2
[ ]
+
F
A
(
v
0). a
1
ln(cos()) + b
1
ln(cos())
2
[ ] +
F
A
(
v
0)
2
. (a
1
a
0
)ln(cos()) + (b
1
b
0
) ln(cos())
2
[ ]

1
]
1
1
1
(05)
with a
0
=1.3347, b
0
=0.57757, a
1
=-1.479, b
1
=-1.5275
However, it has to be pointed out that if the approximations (04 and 05) enable to take into account
adjacency effect for an arbitrary view angle, they implied uniformity of the background as a
function of azimuth. As contributions of the adjacent pixels for a large view angle don't comply to
the symmetry in azimuth, the 6S results, in case of large view angles, have be interpreted more like
107
a sensitivity test to the problem of adjacency effect rather than an actual way to perform adjacency
effect correction .
Reference:
D. TANRE, M. HERMAN and P.Y. DESCHAMPS, Influence of the background contribution upon
space measurements of ground reflectance, Appl. Opt., 20, p. 3676-3684, 1981.
0.8
Radius (km)
F (r)
P
F (r)
R
0 2 4 6 8 10
0.6
0.4
0.2
0.0
1.0
Figure 1: Environment fucntion at satellite level for Rayleigh and Particules.
108
0
0.2
0.4
0.6
0.8
1
0 2 4 6 8 10
z=0. 5km
z=2km
z=5km
z=7km
E
n
v
i
r
o
n
m
e
n
t

F
u
n
c
t
i
o
n

F
R
(
r
)
Radius (km)
Figure 2: Variation of Rayleigh environment fucntion wih sensor altitude.
0
0.2
0.4
0.6
0.8
1
0 2 4 6 8 10
z = 0. 5km
z=2km
z=5km
z=7km
E
n
v
i
r
o
n
m
e
n
t

F
u
n
c
t
i
o
n

F
p
(
r
)
Radius (km)
Figure 3: Same as figure 2 but for particles.
109
0. 0
0. 1
0. 2
0. 3
0. 4
0. 5
0. 6
0 2 4 6 8 10
v
=0
v
=30
v
=60
fit F
r
(r,
v
=30) (10a)
fit F
r
(r,
v
=60) (10a)
F
r
(
r
)
r [km]
Figure
4a: Environment function for a pure molecular atmosphere (lines) for different view zenith angle
(
v
) compared to approximation used in 6S (symbols)
as a function of the distance to the imaged pixel (r).
0. 0
0. 2
0. 4
0. 6
0. 8
1. 0
0 2 4 6 8 10
v
=0
v
=30
v
=60
fit F
a
(r,
v
=30) (10b)
fit F
a
(r,
v
=60) (10b)
F
a
(
r
)
r [km]
Fi
gure 4b: same as Figure 4a but for aerosol.
110
SUBROUTINE GAUSS
Function: Compute for a given n, the gaussian quadrature (the n gaussian angles and the
their respective weights). The gaussian quadrature is used in numerical integration involving the
cosine of emergent or incident direction zenith angle.
111
SUBROUTINE INTERP
Function: To estimate the different atmospheric functions (
S
,
v
,
S
,
v
), T() and S at any
wavelength from the 10 discret computations.(subroutine DISCOM)
Description: The different atmospheric functions (noted f) have been assumed linear as a
function of optical depth , so the interpolation scheme is written,
f() = A
-
The constants A and are interpolated between 0.4 and 3.7 m and extrapolated for the two
extreme intervals 0.25-0.4 and 3.7-4 m.
The spectral dependances for Rayleigh ( = 4) and aerosols ( = 1) are quite different and we
considered the two types of atmosphere separately .
112
SUBROUTINE ISO
Function: Compute the atmospheric transmission for either a satellite or aircraft observation
as well as the spherical albedo of the atmosphere.
Description: The subroutine performs the computation on the basis of the Sucessive Orders
of Scattering method (see subroutine OS). The transmission is obtained directly by initially setting
the bottom of the atmosphere to a isotropic source of radiation. The spherical albedo is computed by
numerical integration (gaussian quadrature) of the transmission function (see Eq. (01) of
CSALBR.f).
113
SUBROUTINE KERNEL
Function: Compute the values of Legendre polynomials used in the successive order of
scattering method.
114
SUBROUTINE MIE (and EXSCPHASE)
Function: To compute, using the scattering of electromagnetic waves by a homogeneous
isotropic sphere, the physical properties of particles whose sizes are comparable to or larger than the
wavelength, and to generate mixture of dry particles.
Description: The interaction of an electromagnetic wave with a absorbing sphere is
described and expressed by the Mie theory (Mie, 1908). This theory has been particularly discussed
by Van de Hulst (Van de Hulst, 1981) and also in part by many other authors (as example Aden,
1951; Deirmendjian et al., 1961; Wyatt, 1962; Kattawar and Plass, 1967; Dave, 1969; Hansen and
Travis, 1974; Liou, 1980). Here, we outline the basic equations of the Mie scattering behind the
computation procedures.
1. Mie Scattering
Let represents the wavelength, r the radius of the sphere, x the Mie's parameter ( x 2 r ), m
the complex index of refraction ( m n
r
in
i
), and the direction of scattered radiation measured
from the forward direction. From the Maxwell's equations, we can defined two complex functions
S
1
(x,m, ) and S
2
(x,m, ) related to the amplitude of the scattered radiation, respectively,
perpendicular and parallel to the plane of scattering
S
1
(x,m, )
(2n 1)
n(n 1)
n 1
a
n
(x,m)
n
(cos ) b
n
(x,m)
n
(cos )
and
S
2
(x,m, )
(2n 1)
n(n 1)
n 1
a
n
(x,m)
n
(cos ) b
n
(x,m)
n
(cos )
1.1 Computation of a
n
(x,m) and b
n
(x,m)
The complex functions a
n
(x,m) and b
n
(x,m) are given by
a
n
(x,m)
n
'
(mx)
n
(x) m
n
(mx)
n
'
(x)
n
'
(mx)
n
(x) m
n
(mx)
n
'
(x)
and
b
n
(x,m)
m
n
'
(mx)
n
(x)
n
(mx)
n
'
(x)
m
n
'
(mx)
n
(x)
n
(mx)
n
'
(x)
115
where the prime denotes derivative of the function with respect of the argument (x or mx), and
where
n
(z x or mx) and
n
(z x) are the Ricatti-Bessel functions defined by
n
(z)
1
2
z
1
2
J
n
1
2
(z) z j
n
(z)
n
(z)
1
2
z
1
2
N
n
1
2
(z) -z n
n
(z)
n
(z)
1
2
z
1
2
H
n
1
2
(2)
(z) z h
n
(2)
(z)
n
(z)+i
n
(z)
where J
n
1
2
, N
n
1
2
and H
n
1
2
(2)
are respectively the Bessel functions of first, second, and third kind,
and where j
n
, n
n
and h
n
(2)
are the corresponding spherical Bessel functions. N
n
1
2
is also called the
Neumann functions and H
n
1
2
(2)
the half integral order Hankel function of the second kind.
In order to make the computational work more convenient, it is useful to introduce the logarithmic
derivative of the Ricatti-Bessel functions (Infeld, 1947; Aden, 1951; Kattawar and Plass, 1967)
D
n
(z)
d
dz
ln
n
(z)
G
n
(z)
d
dz
ln
n
(z)
making use of these equations, a
n
(x,m) and b
n
(x,m) may be rewritten
a
n
(x,m)
n
(x)
n
(x)
D
n
(mx) mD
n
(x)
D
n
(mx) mG
n
(x)
and
b
n
(x,m)
n
(x)
n
(x)
mD
n
(mx) D
n
(x)
mD
n
(mx) G
n
(x)
Expressions of a
n
(x,m) and b
n
(x,m) are now reduced to a ratio of Ricatti-Bessel functions involving
real arguments and a ratio of "D
n
(mx or x) and G
n
(x) " functions which are easily computable. We
reported Figures 1 and 2, examples of a
n
(x,m) and b
n
(x,m) for m = 1.33 - i 0.001 and for x=10 and
x=50 (which means respectively r 0.8 m and r 4.0m at 0.50 m).
Also in order to save time, we use in 6S the criterion defined by Deirmendjian et al., 1961:
"the quantities a
n
and b
n
are terminated either when (a
n
a
n
*
+ b
n
b
n
*
)/ n < 10
-14
"
1.1.1 Computation of the Ricatti-Bessel function.
The ratio of Ricatti-Bessel functions can be reduced to a ratio of spherical Bessel functions with a
real argument x as follow
n
(x)
n
(x)
j
n
(x)
h
n
(2)
(x)
j
n
(x)
j
n
(x) in
n
(x)
The spherical Bessel functions j
n
(x) , n
n
(x) or h
n
(2)
(x) , have different behaviors following they are
below or above the transition line defined by x
2
n(n + 1). Below the transition line
116
(n(n + 1) < x
2
), they behave as oscillating functions of both order and argument, whereas the
behavior becomes monotonic above the transition line (n(n 1) x
2
).
It has been shown by many authors that n
n
(x) or h
n
(2)
(x) can be processed using an upward
recurrence (what ever values of n and x). Functions n
n
(x) are computed using
n
n 1
(x) =
2n+1
x
n
n
(x) + n
n 1
(x)
with n
0
(x)
cos(x)
x
n
1
(x)
cos(x)
x
2
sin(x)
x
Figures 03 show examples of the n
n
(x) function for x=10 and x=50.
For j
n
(x) , we use a similar recurrence
j
n+ 1
(x) =
2n+1
x
j
n
(x) + j
n 1
(x)
but, has it is explained in the paper of Corbat and Uretsky, 1959, the function j
n
(x) cannot be
computed by an upward recurrence "since upward recursion (except in the region of the x-n plane
where j
n
oscillate) would bring about a rapid loss of accuracy". Then, a downward recurrence is
called for, but we have to define the starting value of n, and for that purpose we use the work of
Corbato and Uretsky which is summurized hereafter. Let N be the starting order of the recursion
with N(N 1) x
2
, in their paper, they show "that rather than accurately evaluate j
N
(x) and
j
N 1
(x) to start the process, a very approximately starting the recursion at a higher order will give
a set of numbers which are accurately proportional to the j
n
over the desired range of n from 0 to
N". Let j
n
be one of these numbers.
They propose to define the higher order by
N
'
ln
N
ln2
A +
B u
'
(2-u
' 2
)
2(1-u
' 2
)

1
]
1
where A=0.10 and B=0.35,
N
2
30
(this value comes from the fact that generally computers can store floating
points numbers with a 30 binary digit mantissa),
and u
'
2x (2N
'
+ 1) with N' =N or N'= x
1
2
+
ln
N
ln2
Bx such that be
the lower, with however N
'
N.
To avoid computational difficulties above the transition line, Corbato and Uretsky worked with the
ratio r
n
j
n+1
j
n
using the recurrence relation
r
n-1
x
2n + 1 xr
n
117
with the starting condition r 0 . The recursion is continued downward until a ratio r
n
which
exceeds unity is reached. Then, they set j
n+ 1
r
n
and j
n
1, and continue downward using the
recurrence relation
j
n 1
(x) =
2n+1
x
j
n
(x) + j
n+ 1
(x)
The positive number j
n
is defined by j
n
(x) j
n
(x) with a constant of proportionality obtained
from the relation
=( j
0
(x)- x j
1
(x))cos(x) + x j
0
(x)sin(x)
Figures 03 also show examples of the j
n
(x) function for x=10 and x=50.
1.1.2 Computation of the D
n
(mx or x) and G
n
(x) function.
As Kattawar and Plass, 1967, have pointed out, the procedure of computing D
n
(z) by an upward
recurrence is unstable, then a downward process is needed, and D
n
(z) is defined using
D
n 1
(z) n z
1
D
n
(z) n z
Calculations have to be started at an order n
'
>> z with a starting value which is not really
important because the serie converges rapidly to the exact value (then D ' (z) 0 is a convenient
value). When n z , D
n
(z) becomes oscillatory, and then there is no problem for the calculation in
using the recurrence relation. For practical reasons, we selected in 6S
'
as defined for j
n
.
Kattawar and plass have also shown that G
n
(x) may always be calculated using an upward process
with a starting value G
0
(x) i
G
n
(x) n x
1
G
n1
(x) n x
Figure 04 reportes examples of the D
n
(x) function, Figures 05 the D
n
(mx) function, and Figures 06
theG
n
(x) function.
1.2 Computation of
n
(cos ) and
n
(cos )
Functions
n
and
n
depend of only the scattering angle . They are related to the associated
Legendre polynomials P
n
1
(cos )
n
(cos )
1
sin
P
n
1
(cos )
n
(cos )
d
d
P
n
1
(cos )
and are computed from upward recurrence relations defined as follow
n
n+1
(cos ) (2n+1) cos
n
(cos ) - (n+1)
n-1
(cos )
n+1
(cos ) (n+1) cos
n+1
(cos ) - (n+2)
n
(cos )
118
with the starting values
0
(cos ) 0
1
(cos ) 1
Examples of functions
n
and
n
are shown Figures 07 for n=1 to 6 and for 0<<90.
2. Computation of the physical properties of a particle (see for example Liou, 1980).
2.1 Extinction
The extinction cross section
e
, which denotes the amount of energy removed (scattered and
absorbed) from the original beam by the particle, is obtained considering a point in the forward
direction ( =0) in the "far field". If we consider an isotropic homogeneous sphere, the extinction
cross section is given by
e
( ,r,m)
4
2 ( )
2
e
S(x,m, 0) [ ]
with
S(x,m, 0) S
1
(x,m, 0) S
2
(x,m, 0)
1
2
(2n 1)[a
n
(x,m) b
n
(x,m)
n 1
]
Thus the extinction efficiency Q
e
is defined by
Q
e
( ,r,m)
e
( ,r,m)
r
2
2
x
2
(2n 1)
e
[a
n
(x,m) b
n
(x,m)
n 1
]
2.2 Scattering
The scattering cross section
s
is derived by a similar way, but considering a scattered light in an
arbitrary direction, by
s
( ,r,m)
x r
2
[S
1
(x,m, )S
1
*
(x,m, ) S
2
(x,m, )S
2
*
(x,m, )]sin d
0
Owing of the functions
n
and
n
, we have to integrate products of the associated Legendre
polynomials. Using the orthogonal and recurrence properties of these polynomials, the scattering
cross section can be written
s
( ,r,m)
2
x r
2
(2n 1)[a
n
(x,m)a
n
*
(x,m) b
n
(x,m)b
n
*
(x,m)]
n 1
where the asterisk denotes the complex conjugate value, and the scattering efficiency Q
s
can be
evaluated by the relation
Q
s
( ,r,m)
s
( ,r,m)
r
2
2
x
2
(2n 1)[a
n
(x,m)a
n
*
(x,m) b
n
(x,m)b
n
*
(x,m)]
n 1
2.3 Absorption
The absorption cross section
a
and the absorption efficiency Q
a
can be deduced from
119
a
( ,r,m)=
e
( ,r,m)
s
( ,r,m) Q
a
( ,r,m) Q
e
( , r, m)-Q
s
( ,r,m)
2.4 Phase function
On the basis of the Stokes parameters, the intensity I of the electromagnetic waves at each point and
in any given direction can be related to the incident intensity I
0
by
I =M
11
I
0
with
M
11
( ,r,m, )=
1
2x
2
[S
1
(x,m, )S
1
*
(x,m, )+S
2
(x,m, )S
2
*
(x,m, )]
The angular distribution of the scattered energy for a single sphere (also called Phase function)
P
11
( , r,m, ) can be defined by
M
11
( , r,m, ) =
s
( ,r,m)
4 r
2
P
11
( , r,m, )
then
P
11
( ,r,m, )=
2
(x r)
2
s
( ,r,m)
[S
1
(x,m, )S
1
*
(x,m, )+S
2
(x,m, )S
2
*
(x,m, )]
It can be check that
P
11
( ,r,m, )
0
0
2
sin d d 4
3. Physical properties of a sample of identical particles
We now consider a sample of identical particles whose size is described by the size distribution
n(r) (in cm
-3
m
-1
) such that
n(r) dr
0
=
dN(r)
dr
dr
0
=1
where d N(r) represents the number of particle per unit volume having a radius between r and
r+dr.
In 6S, we selected several possibilities to represent the size distribution, thus the user will be
allowed to choice between 4 options:
1- a Junge power-law function . Junge, 1952, showed that the size distribution of aerosols whose
radii are larger than 0.1m may be described by
dN(r)
dlogr
ln(10) c r
0

1
r

_
,
1
or
dN(r)
dr
c r
0

1
r

_
,
with varying between 3 and 5, c the number density of particles with radius r
0
and r
0
an
arbitrary radius.
Figure 08-a shows an example of Junge Power-Law function which is the "Model C"
defined by Deirmendjian, 1969, for c.r
0
=1 and =4
120
2- a Modified Gamma distribution function. Used by Deirmendjian, 1964, to compute scattering
properties of water clouds and haze and to fit aerosol measurements. Also employed by Mie
in the Mie and diffraction calculations.
dN(r)
dr
A r r
0
( ) exp -b r r
0
( )
( )
with r
0
=1m
An example of Modified Gamma distribution function is given Figure 08-b (Volcanic Ash
defined in WCP 112, A=5461.33, =1.0, =0.5, b=16).
3- a Log-Normal distribution function . Based on the Junge power-law function, Davies, 1974,
introduced this function to take into account large particles.
dN(r)
dlogr
N
2 log
exp -
1
2
log r -log r
M
log

_
,
2

_
,

where r
M
is the mean radius of the particle, and the standard deviation of r.
We reported Figure 08-c examples of Log-Normal distribution functions which are the 3
three components of the "Continental Model" defined in WCP 112 (see AEROSO to find r
M
and ).
4- sun photometer measurements . You enter directly d V(r)/ dlogr r
4
dN(r) / dr .
The Figure 09 shows the same function than Figures 08 but for dV(r)/dlog(r).
Under the assumption of "independent scattering" which means that particles are sufficiently far
from each other compared to the incident wavelength to consider just one scattering, it is possible to
add scattered intensities independently of the phase of the wave. Then we can defined the radiative
characteristics upon the particle size distribution by
The extinction (e), scattering (s) and absorption (a) coefficient
k
e,s,a
( ,m) Q
e,s,a
( ,r,m) r
2
rmin
r
max
dN(r)
dr
dr
The normalized phase function
P( ,m, )
1
k
s
( ,m)
M
11
( ,r,m, ) 4
rmin
r
max
r
2
dN(r)
dr
dr
We now introduce the single scattering albedo
0
which represents the percentage of
energy removed from the incident beam which will reappear as a single scattered
radiation.
0
( , m)
k
s
( , m)
k
e
( , m)
121
Computationally, k
e,s,a
( ,m) and P( ,m, )are integrated step by step following:
k
e,s,a
( ,m) Q
e,s,a
( ,r,m) r
2
r
min
r
max
dN(r)
dr
r
and
P( ,m, )
1
k
s
( ,m)
M
11
( ,r,m, ) 4
r
min
r
max
r
2
dN(r)
dr
r ,
where r is defined by
log
r + r
r

_
,
0. 03
The value 0.03 has been selected in order to preserve a good accuracy with a reasonable
computational time. For example D'Almeida used a very small step width, 0.011, for the
computations given in his book (D'Almeida et al., 1991). The logarithmic expression of r comes
from the fact that size distributions can be frequently described by a logarithmic shape (Junge,
1952; Davies, 1974).
Finally, and, in order to save computational time, we defined a criterion on the summation such that
the computations are not performed either when
n
i
n
r
2
dN(r)
dr
r

_
,
1
< 10
8
where n
i
n is the percentage density of particles (cf. subroutine AEROSO for some examples). The
latter criterion has been tested between 0.4 and 4.0 m.
4. Physical properties of a mixture of aerosol type
We now consider a mixture of particles originating from different sources (4 max.). The mixing is
treated in the same way that the one used to generated the data base in the AEROSO subroutine.
Let us recall that the mixture of individual components (or type) of an aerosol is characterized by
the percentage density of particles n
i
n , and if we assume that the particles are spherical, each type
i is described by its size distribution (then by its microphysical identity: r
M i
and
i
see Table 1 for
some examples (Shettle and Fenn, 1976; World Climate Programme, 1986), and by its complex
refractive index m
i
(see Table 2, from (Shettle and Fenn, 1976; Shettle and Fenn, 1979; World
Climate Programme, 1986; D'Almeida et al., 1991). For the size distribution, the Log-Normal
distribution is well adapted to emphasize the individual components of a mixture (Davies, 1974,
D'Almeida et al., 1991).
122
5. Examples and comparisons
The comparison of the computed normalized at 550 nm K
e
, K
s
and
0
values that obtained by 6S
with those given by World Climate Programme, 1986 are reported for a Continental and an Urban
dry aerosol model respectively Table 3 and Table 4. Also reported in theses tables, are the
asymmetry parameters g. In 6S, this parameter is already computed in an another subroutine, but we
can compute it here using
g
cos P( ,m, )
1
+1
dcos
P( ,m, )
1
+1
dcos
The comparison of the phase function of a Continental model (WMO/WCP-112) computed by the
MIE subroutine with those by a precise code (AEROSO subroutine) is reported Figure 10 for
several wavelengths. Also, we show, Figure 11, the phase function computed using the volumic
distribution dV/dlogr provided by a CIMEL sunphotometer during the SCAR-A field experiment
(Sulfate Clouds Aerosol and Reflectances - America) that took place in July 1993 in the Eastern
USA.
123
The parameters are:
For IAER=8 Multimodal Log Normal (up to 4 modes)
r
min
,r
max
,icp
then for k=1 to icp, enter:
, r
M
, C
ij
r
n
(
j
), j=1,10
r
i
(
j
), j=1,10
where r
min
and r
max
are the radii min and max of the aerosol
icp the number of mode (component)
and r
M
are parameters of the Log-Normale size distributions
Cij is the percentage density of particles (see SUBROUTINE AEROSO)
r
n
and r
i
are the real and imaginary index of refraction of each component
with r=r
n
-i r
i
. You have to enter these parameters for the 10 wavelengths
used to compute the atmospheric signalwhich are:
0.400, 0.488, 0.515, 0.550, 0.633, 0.694, 0.860, 1.536, 2.250, 3.750
For IAER=9 Modified gamma distribution
r
min
,r
max
, b,
r
n
(
j
), j=1,10
r
i
(
j
), j=1,10
where , b, and are the parameters of the Modified Gamma size distribution
For IAER=10 Junge Power-Law distribution
r
min
,r
max
r
n
(
j
), j=1,10
r
i
(
j
), j=1,10
where is the parameter of the Junge Power-Law size distribution
For IAER=11 Sun Photometer distribution (50 values max)
irsunph
for k=1 to irsunph enter: r and dV/dlogr
r
n
(
j
), j=1,10
r
i
(
j
), j=1,10
where irsunph is the number of value and dV/dlogr is usually profided by sunphotometers.
124
Table 1: Microphysical characteristics of the aerosol type (dry particles)
used for the comparisons shown Tables 3 and 4 (from WMO-WCP112).
Dust-Like Water Soluble Oceanic Soot
r
Mi
(m) 0.500 0.0050 0.30 0.0118
i 2.990 2.990 2.51 2.00

Table 2: Complexe refractive indexes of the aerosol types (dry particles)
used for the comparisons shown Tables 3 and 4 (from WMO-WCP112).
Dust-Like Water Soluble Oceanic Soot
(m) n
r
n
i
n
r
n
i
n
r
n
i
n
r
n
i
0.400 1.530 8.00E-3 1.530 5.00E-3 1.385 9.90E-9 1.750 0.460
0.488 1.530 8.00E-3 1.530 5.00E-3 1.382 6.41E-9 1.750 0.450
0.515 1.530 8.00E-3 1.530 5.00E-3 1.381 3.70E-9 1.750 0.450
0.550 1.530 8.00E-3 1.530 6.00E-3 1.381 4.26E-9 1.750 0.440
0.633 1.530 8.00E-3 1.530 6.00E-3 1.377 1.62E-8 1.750 0.430
0.694 1.530 8.00E-3 1.530 7.00E-3 1.376 5.04E-8 1.750 0.430
0.860 1.520 8.00E-3 1.520 1.20E-2 1.372 1.09E-6 1.750 0.430
1.536 1.400 8.00E-3 1.510 2.30E-2 1.359 2.43E-4 1.770 0.460
2.250 1.220 9.00E-3 1.420 1.00E-2 1.334 8.50E-4 1.810 0.500
3.750 1.270 1.10E-2 1.452 4.00E-3 1.398 2.90E-3 1.900 0.570
125
Table 3: Comparison between 6S and WMO-WCP112
for a Continental model (normalized value-dry particles)
K
ext
K
sca
0 g
6S WMO 6S WMO 6S WMO 6S WMO
0.400 1.40 1.40 1.27 1.27 0.902 0.901 0.643 0.646
0.488 1.14 1.14 1.03 1.03 0.900 0.898 0.637 0.640
0.515 1.08 1.08 0.967 0.967 0.899 0.897 0.635 0.638
0.550 1.00 1.00 0.893 0.891 0.893 0.891 0.634 0.637
0.633 0.849 0.849 0.755 0.754 0.890 0.888 0.629 0.633
0.694 0.760 0.760 0.671 0.669 0.881 0.879 0.628 0.631
0.860 0.577 0.577 0.487 0.486 0.844 0.841 0.629 0.633
1.536 0.282 0.283 0.212 0.212 0.753 0.750 0.641 0.645
2.250 0.150 0.151 0.115 0.115 0.765 0.761 0.738 0.741
3.750 0.101 0.103 0.0796 0.0805 0.790 0.785 0.777 0.779
Table 4: Comparison between 6S and WMO-WCP112
for an urban model (normalized value-dry particles)
K
ext
K
sca
0 g
6S WMO 6S WMO 6S WMO 6S WMO
0.400 1.48 1.48 0.980 0.976 0.664 0.660 0.600 0.600
0.488 1.16 1.17 0.766 0.762 0.658 0.654 0.594 0.593
0.515 1.09 1.09 0.715 0.711 0.655 0.651 0.592 0.592
0.550 1.00 1.00 0.651 0.647 0.651 0.647 0.591 0.591
0.633 0.828 0.829 0.535 0.532 0.646 0.641 0.587 0.587
0.694 0.733 0.733 0.466 0.462 0.635 0.631 0.585 0.585
0.860 0.542 0.542 0.322 0.319 0.593 0.588 0.584 0.583
1.536 0.242 0.243 0.111 0.111 0.460 0.455 0.564 0.565
2.250 0.123 0.124 0.0428 0.0426 0.347 0.342 0.583 0.585
3.750 0.0647 0.0659 0.0177 0.0181 0.274 0.274 0.579 0.587
126
Figures 01: Examples of a a
n
(m,x) function.
0. 0
0. 20
0. 40
0. 60
0. 80
1. 0
1. 2
0 10 20 30 40 50 60 70
Re[a
n
(m,x=10)]
Re[a
n
(m,x=50)]
R
e
[
a
n
(
m
,
x
)
]
Order n
m = 1.33 - i 0.001
-0. 60
-0. 40
-0. 20
0. 0
0. 20
0. 40
0. 60
0. 80
0 10 20 30 40 50 60 70
Im[a
n
(m,x=10)]
Im[a
n
(m,x=50)]
I
m
[
a
n
(
m
,
x
)
]
Order n
m = 1.33 - i 0.001
Figures 02: Examples of a b
n
(m,x) function.
0. 0
0. 20
0. 40
0. 60
0. 80
1. 0
1. 2
0 10 20 30 40 50 60 70
Re[b
n
(m,x=10)]
Re[b
n
(m,x=50)]
R
e
[
b
n
(
m
,
x
)
]
Order n
m = 1.33 - i 0.001
-0. 60
-0. 40
-0. 20
0. 0
0. 20
0. 40
0. 60
0. 80
0 10 20 30 40 50 60 70
Im[b
n
(m,x=10)]
Im[b
n
(m,x=50)]
I
m
[
b
n
(
m
,
x
)
]
Order n
m = 1.33 - i 0.001
Figures 03: Examples of spherical Bessel functions j
n
(x) and n
n
(x)
-0. 15
-0. 10
-0. 05
0. 00
0. 05
0. 10
0. 15
0 10 20 30 40 50 60 70
j
n
(x=10)
j
n
(x=50)
j
n
(
x
)
Order n
-0. 20
-0. 15
-0. 10
-0. 05
0. 00
0. 05
0. 10
0. 15
0. 20
0 10 20 30 40 50 60 70
n
n
(x=10)
n
n
(x=50)
n
n
(
x
)
Order n
127
Figure 04: Examples of a D
n
(x) function.
-20.00
-15.00
-10.00
-5. 00
0. 00
5. 00
10. 00
15. 00
20. 00
0 20 40 60 80 100
Re [D
n
(x=10)]
Re [D
n
(x=50)]
R
e

[
D
n
(
x
)
]
Order n
Figures 05: Examples of a D
n
(mx) function.
-10.00
-5. 00
0. 00
5. 00
10. 00
0 20 40 60 80 100
Re [D
n
(m,x=10)]
Re [D
n
(m,x=50)]
R
e

[
D
n
(
m
x
)
]
Order n
m=1.35 - i 0.001
10
- 3
10
- 2
10
- 1
10
0
10
1
10
2
0 20 40 60 80 100
Im [D
n
(m,x=10)]
Im [D
n
(m,x=50)]
I
m

[
D
n
(
m
x
)
]
Order n
m=1.35 - i 0.001
Figures 06: Examples of a G
n
(x) function.
-1. 40
-1. 20
-1. 00
-0. 80
-0. 60
-0. 40
-0. 20
0. 00
0 20 40 60 80 100
Re[ G
n
(x=10)]
Re[ G
n
(x=50)]
R
e
[
G
n
(
x
)
]
Order n
-1. 20
-1. 00
-0. 80
-0. 60
-0. 40
-0. 20
0. 00
0 20 40 60 80 100
I m[ G
n
(x=10)]
I m[ G
n
(x=50)]
I
m
[
G
n
(
x
)
]
Order n
128
Figures 07: Examples of functions
n
() and
n
() for n=1 to 6.
- 5
0
5
10
15
20
25
0 30 60 90
n
Scattering Angle
1
2
3
4
5
6
-20
-10
0
10
20
0 30 60 90
n
Scattering Angle
1
2
3
4
5
6
Figure 08-a: Junge Power-law function: Model C (Deirmendjian, 1954)
10
-10
10
-7
10
-4
10
-1
10
2
10
5
10
-4
10
-3
10
-2
10
-1
10
0
10
1
10
2
d
N
(
r
)
/
d
r
r (m)
Model C (Deirmendjian 1969)
Figure 08-b: Modified Gamma distribution function: Volcanic Ash (WCP 112)
10
-10
10
-7
10
-4
10
-1
10
2
10
5
10
-4
10
-3
10
-2
10
-1
10
0
10
1
10
2
d
N
(
r
)
/
d
r
r (m)
Volcanic Ash
(WMO/WCP 112)
129
Figure 08-c: Log-Normal distribution function: Continental Model (WCP 112)
10
-10
10
-7
10
-4
10
-1
10
2
10
5
10
-4
10
-3
10
-2
10
-1
10
0
10
1
10
2
d
N
(
r
)
/
d
l
o
g
r
r (m)
Dust Like
Soot
Water Soluble
Continental Model
(WMO/WCP 112)
Figure 09: Same as Figures 8 but represented for dV/dlogr
10
- 6
10
- 5
10
- 4
10
- 3
10
- 2
10
- 1
10
0
10
1
10
- 3
10
- 2
10
- 1
10
0
10
1
10
2
Model C (Deirmendjian 1954)
Volcanic Ash (WMO/WCP 112)
Continental Model (WMO/WCP 112)
d
V

/

d
l
o
g
r
r (m)
130
Figure 10: Phase function (dry particle) as computed by the MIE subroutine and
by the one generated by AEROSO subroutine (exact case).
0.1
1
10
100
0 30 60 90 120 150 180
Continental Model
From AEROSO (Exact Case)
P
h
a
s
e

f
u
n
c
t
i
o
n
Scattering angle
3.750m
0.860m
0.440m
From MIE
Figure 11: Phase function as computed by MIE subroutine using the dV/dlogr provided
by a sunphotometer CIMEL during the SCAR-A experiment (Hog Island, July 11, 1993).
0.1
1
10
100
0 30 60 90 120 150 180
441 nm
873 nm
P
h
a
s
e

f
u
n
c
t
i
o
n
Scattering Angle
131
REFERENCES
A.L. ADEN, Electromagnetic Scattering from Spheres with Sizes Comparable to the Wavelength. J.
Appl. Phys., 22, 5, 601-605, 1951
F.J. CORBATO and J.L. URETSKY, Generation of Spherial Bessel Functions in Digital
Computers. J. Assoc. Computing Machinery, 6, 366-375, 1959
G.A. D'ALMEIDA, P. KOEPKE, and E.P. SHETTLE, Atmospheric Aerosols Global Climatology
and Radiative Characteristics, A. Deepak Publishing, Hampton, 1991.
J.V. DAVE, Scattering of Visible Light by Large Water Spheres. Appl. Opt., 8, 1, 155-164, 1969
C.N. DAVIES, Size distribution of atmospheric particles. J. Aerosol Sci., 5, 293-300, 1974
D. DEIRMENDJIAN, Scattering and Polarization properties of water clouds and hazes in the
visible and infrared. Appl. Opt., 3, 187-196, 1964
D. DEIRMENDJIAN, R. CLASEN, and W. VIEZEE, Mie Scattering with Complex Index of
Refraction. J. Opt. Soc. Am., 51, 6, 620-633, 1961
D. DEIRMENDJIAN, Electromagnetic scattering on spherical polydispersion. Elsevier Ed. New-
York, 1969
J.E. HANSEN and L. TRAVIS, Light scattering in planetary atmospheres. Space Sci. Rev., 16, 527-
610, 1974
L. INFELD, The Influence of the Width of the Gap upon the Theory of Antennas. Quart. Appl.
Math., 5, 2, 113-132, 1947
C.E. JUNGE, Gesetzmigkeiten in der Droenverteilung atmosphrischer Aerosole ber dem
Kontinent. Ber d. Deirsch Wetterdienst U.S.-Zone, 35, 261-277, 1952
G.W. KATTAWAR and G.N. PLASS, Electromagnetic Scattering from Absorbing Spheres. Appl.
Opt., 6, 8, 1377-1382, 1967
K.N. LIOU, An Introduction to Atmospheric Radiation, Academic Press Inc., San Diego, 1980.
G. MIE, Beigrade zur Optik trber Medien, speziell kolloidaler Metallsungen. Ann. Physik., 25,
377-445, 1908
E.P. SHETTLE and R.W. FENN, Models of atmospheric aerosols and their optical properties, in:
Optical Properties in the Atmosphere, AGARD-Cp-183, NTIS, ADA 028615, 1976.
E.P. SHETTLE and R.W. FENN, Models for the aerosol of the lower atmosphere and the effect of
humidity variations on their optical properties, AFGL-TR-79-0214, Environmental Research
Paper No 675, NTIS, ADA 085951, 1979.
H.C. VAN DE HULST, Light Scattering by Small Particles, Dover Publications, New York, Dover,
1981.
132
WORLD CLIMATE PROGRAMME, WCP-55, Report of the expert meeting on aerosols and their
climatic effects (Eds. A. Deepak and H.E. Gerber), World Meteorological Organization,
Geneva, 1983.
WORLD CLIMATE PROGRAMME, WCP-112, A preliminary cloudless standard atmosphere for
radiation computation, World Meteorological Organization, WMO/TD-No 24, Geneva, 1986.
P.J. WYATT, Scattering of Electromagnetic Plane Waves from Inhomogeneous Spherically
Symmetric Objects. Phys. Rev., 127, 5, 1837-1843, 1962
133
SUBROUTINE ODA550
Function: To compute the extinction cross section and the aerosol optical depth at = 550
nm from the vertical distributions of the particle density (in particules/cm
3
).
Description: We have considered the 2 profiles, suggested by Mc Clatchey et al (1971),
corresponding to a visibility of 23 (clear) and 5 km (hazy) at ground level. The total numbers of
aerosols for the clear atmosphere have been adjusted so that the total extinction coefficient at =
550 nm becomes identical to the values used by Elterman (1964).
This total extinction coefficient K (in km
-1
) is obtained from
K
550
(z)
550
10
3
N(z)
where is the extinction cross section in m
2
and N(z) the particules density (in part/cm
3
)
(the factor 10
-3
is to obtain an extinction coefficient in km
-1
). was computed with the same
aerosol model as the one defined by Mc Clatchey, index of refraction equal to 1.50 and size
distribution similar to Deirmendjian's model "C" (1969) (cut off has been extended from 5 to 10
m). The computed value of
550
is 0.056032.
The optical thickness is defined by
550
K
550
(z) dz
0
+
We obtain the optical thicknesses at 550 nm, 0.235 and 0.780 respectively for the two
standard visibilities 23 and 5km. For another visibility, we compute a new profile particle density
from those defined for 23 and 5km. The calculations were made using the following interpolations,
N(z)
a(z)
VIS
+ b(z)
We obtain for example :
134
= 0.152 for V = 50km,
= 0.520 for V = 8km.
Reference :
R.A. Mc CLATCHEY, R.W. FENN, J.E.A. SELBY, F.E. VOLZ, J.S. GARING. Optical properties
of the Atmosphere (revised), AFCRL 71-0279, Env. Research Paper 354, Bedford,
Massachusetts, U.S.A., 1971.
D. DEIRMENDJIAN, Electromagnetic Scattering on Spherical Polydispersions , 1969.
L. ELTERMAN , Rayleigh and Extinction Coefficient to 50km for the region 0.27m to 0.55m.
Applied Optics, 10, 1139-1145, 1964.
135
SUBROUTINE ODRAYL
Function: To compute the molecular optical depth as a function of wavelength for any
atmosphere defined by the pressure and temperature profiles.
Description: The optical depth is written
(z) dz
0
+
where
(z) is the molecular extinction coefficient at altitude z and for wavelength . It can
be obtained from
(z)
N
r
(z)10
5
,
with N
r
(z) is the molecules number/cm
3
at altitude z, and
the extinction (or scattering)

cross section in cm
2
.
These two quantities are defined by
8
3
(n
s
2
1)
2
3
4
N
s
2
6 + 3
6 7
and
N
r
(z) N
s
P(z)
1013.25
273.15
T(z)
where P(z) and T(z) are respectively the pressure and the temperature at the altitude z. Recall
that n
S
is the air refractive index, N
S
the molecular density at z=0 in STP conditions, and the
molecular depolarization factor.
We have taken:
* for refractive index
(n
s
1).10
8
8342.13 +
2406030
130
2
+
15997
38.9
2
where is the frequency in cm
-1
136
* N
S
= 2.54143.10
19
* and the depolarization factor = 0.0279 following Young's (1980).
This depolarization factor is also used to compute the Rayleigh phase function (see routine
CHAND.f) according to:
P()
3
4
1
1 + 2
(1+ cos
2
) +
3
1 + 2
where is the scattering angle, and = /(2-).
References :
B. EDLEN, The refractive Index of Air. Meteorologia, 2, 71-80, 1966
L. ETERMAN, Rayleigh and Extinction Coefficients to 50 km for the Region 0.27 m to 0.55 m
Appl. Opt., 10, p. 1139-1145, 1964
D.V. HOYT, A redetermination of the Rayleigh Optical Depth and its Application to selected Solar
Radiation Problems, J. Appl. Meteor, 16, p. 432-436, 1977.
A. T. YOUNG, Revised Depolarization Corrections for Atmospheric Extinction. Applied Optics,
19, 3427-3428, 1980
137
SUBROUTINE OS
Function: Compute the atmospheric intrinsic reflectance for the case of eihter satellite or
aircraft observation. Also compute the downward radiation field needed by the integral formula of
and ' (see Chapter I, 2.5.1, Eqs. 25 and 26) used in the computation in case of a non
lambertian target.
Description: The general purpose of the successive order of scattering is to solve
numericaly the equation of radiative transfer for upward (Eq. 01.) and downward radiation (Eq. 02.)
at any optical thickness . If
1
is the total optical thickness and the cosine of the view angle, then
we can write:
I(;, ) I(
1
;, )
(
1
)/
e + J(' ; , )
(' )/
e
d'
(1 0) (01)
I(; , ) I(0; , )
/
e + J( ' ;, )
(' )/
e
d'
(1 0) (02)
where the source function, J(;,) accounts for the interaction of the present radiation field with
the particles of the layer located at , so that:
J( ;, )
0
4
I(;' , ' ) P(, ;', ' ) d' d'
1
1
0
2
0
4
F
0
P(, ;
0
,
0
)
/
0 e (03)
The second term of equation (03) represents the sun source F
0
transversing the path along
(
0
,
0
) directly to the level and then being scattered in direction (,) (primary scattering).
To solve this differential equation, one has to fix boundary conditions which are:
I(0;,) =0 (04)
I(
1;
,) = 0 (05)
These express the fact that there is no diffuse downward nor upward radiation at the top and the
bottom of the finite atmosphere.
138
The convention is to describe the atmosphere with the top at =0 and the bottom at =
1
. The
upward radiation correspond to + and the downward to with (1>0) as depicted by figure 1.
Figure 1: Schematic view of the radiative transfer problem for a plane parralel atmosphere.
Top
Bottom
= 0
= 0
=
1
I(0;,)
I(
1
;,)
I(;,)
I(
1
;-,)
I(;-,)
The successive order of scattering consists of solving numerically the equation of transfer by
iteration. First, the equation is solved for each layer considering only the primary scattering
radiation (one interaction between the source (sun) and the atmosphere), giving for Eqs. (01) and
(02):
I
(1)
(; , )
0
4
F
0
P(, ;
0
,
0
)
/
0 e (1 0) (06)
I
(1)
(; , )
0
4
F
0
P(, ;
0
,
0
)
/
0 e (1 0) (07)
Then for higher order of scattering we write:
139
I
(n)
(
j
; , )
1
J
(n)
(
j
;, )
(
j
)/
[ ] e
i j
p
(08)
I
(n)
(
j
; , )
1
J
(n)
(
j
;, )
(
j
)/
[ ] e
i 1
j
(09)
where p represents the number of layers used for the decomposition of the atmosphere,
j
the optical
thickness at level j and the increment in optical thickness between two successive layer. J
(n)
is
computed from I
(n-1)
by:
J
(n)
(;, )
0
4
I
(n)
(;', ' )P( , ;' , ' )d' d'
1
1
0
2
(10)
In the code a numerical integration of Eq. 10 is performed using the decomposition in Fourier series
(for ), the legendre polynomial and Gaussian quadrature (for ).
140
SUBROUTINE SCATRA
Function: To compute the scattering transmission functions for the three atmospheric models,
rayleigh, aerosol and a mixture of both on the two paths (downward and upward). We also compute
the spherical albedo.
Description: As in ATMREF.f, we have to compute the transmission function and albedo in
three different cases and three sensor configurations. Again, the accuracy of the computation of 6S
was a concern, so the approximation adopted in 5S has been replaced by using the sucessive order
of scattering method (ISO.f) for the aerosol case and the mixed case, or when the sensor was inside
the atmosphere on board an aircraft. For the Rayleigh atmosphere we used an accurate analytical
formula which has sufficient accuracy and enables us to save computer time. The formula is
explicity coded into SCATRA.f for the transmission and call CSALBR.f for the albedo. For ground
measurements, the upward transmission is set to 1.0 and the spherical albedo to 0.0, because we
neglect the atmosphere between the sensor and the target.
We only give here the formula of the Rayleigh transmission which is based on the two stream
method adapted to the case of a single scattering albedo equal to 1.0 (Rayleigh case). The total
transmission on path of length , T() can be approximated by:
T()
[(2/ 3) +] +[(2/ 3) ]e
(4 / 3) +
R
(01)
where
R
is the Rayleigh optical thickness.
Figure 1 compares the accuracy of Eq. 01 to the "exact" computation (Successive Order of
Scattering).
141
References
J.H. JOSEPH, W.J. WISCOMBE, J.A. WEINMAN, The Delta-Eddington Approximation for
Radiative Flux Transfer,.J. Atmos. Sci. 33, p.2242-2459, 1976.
W.E. MEADOR, W.R. WEAVER, Two-Stream Approximations to Radiative Transfer in Planetary
Atmospheres: a Unified Description of Existing Methods and a New Improvement, J. Atmos.
Sci. 37, p. 630-643, 1980.
R.H.WELCH, W.G. ZDUNKOWSKI, Back Scattering Approximations and their Influence on
Eddington-Type Solar Flux Calculation, Beitr. Phys. Atmosph., 55, no 1, p. 28-42, 1982.
W.G. ZDUNKOWSKI, R.M. WELCH, G. KORB, An Investigation of the Structure of Typical
Two-Stream Methods for the Calculation of Solar Fluxes and Heating Rates in Clouds, Beitr.
Phys. Atmosph.53, no 2, p. 147-166, 1980.
Figure 1: Accuracy of Eq. 01
0.00
0.20
0.40
0.60
0.80
1.00
0.000
0.002
0.004
0.006
0.008
0.010
0 10 20 30 40 50 60 70
exact results
analytical expression
absolute difference
Zenith observation angle ()
T
r
a
n
s
m
i
s
s
i
o
n

f
u
n
c
t
i
o
n

T
(
)
A
b
s
o
l
u
t
e

d
i
f
f
e
r
e
n
c
e
142
SUBROUTINE TRUNCA
Function: decompose the aerosol phase function in series of Legendre polynomial used in
OS.f and ISO.f and compute truncation coefficient f to modify aerosol optical thickness and single
scattering albedo
0
according to:
' (1
0
f)
0
'
0
(1 f)
(1
0
f)
References
J. LENOBLE, Radiative Transfer in Scattering and Absorbing Atmospheres: Standard computional
procedures, pp 83-84, A. Deepak Publishing, 1985
143
DESCRIPTION OF THE SUBROUTINES USED FOR BRDF GROUND
144
145
SUBROUTINE HAPKALBE
Function: To compute the spherical albedo following the the BRDF computed from Hapke
(1981) Model.
Description: The target spherical albedo s is equal to the flux reflected by the target divided
by the incoming flux for an isotropic source. It is defined by the relation:
s=
0
/2
a(
s
)cos(
s
)sin(
s
)d
s
0
/2
cos(
s
)sin(
s
)d
s

where a(
s
) is the directional albedo for a parallel solar beam, and is given by
a(
s
)

s
,
v
,
s
v
( )cos(
v
)sin(
v
)d
v
d
0
/ 2
0
2
cos(
v
)sin(
v
)d
v
d
0
/ 2
0
2
with (
s
,
v
,), the bidirectionnal reflectances generated by the users's inputs (see
HAPKBRDF).
146
SUBROUTINE IAPIALBE
Function: Same as HAPKALBE but for a BRDF from the subroutine IAPIBRDF.
SUBROUTINE MINNALBE
Function: Same as HAPKALBE but for a BRDF from the subroutine MINNBRDF.
SUBROUTINE OCEALBE (and GLITALBE)
Function: Same as HAPKALBE but for a BRDF from the subroutine OCEABRDF.
SUBROUTINE RAHMALBE
Function: Same as HAPKALBE but for a BRDF from the subroutine RAHMBRDF.
SUBROUTINE ROUJALBE
Function: Same as HAPKALBE but for a BRDF from the subroutine ROUJBRDF.
SUBROUTINE VERSALBE
Function: Same as HAPKALBE but for a BRDF from the subroutine VERSBRDF.
SUBROUTINE WALTALBE
Function: Same as HAPKALBE but for a BRDF from the subroutine WALTBRDF.
147
SUBROUTINE BRDFGRID
Function: To generate a BRDF following user's inputs.
Description: The user enters the value of for a sun at a given zenith sun angle
s
by steps of
10 for zenith view angles
v
(from 0 to 80 and the value at 85) and by steps of 30 for azimuth
view angles
v
from 0 to 360. The user does the same for a sun which would be at
v
. In addition,
the spherical albedo of the surface has to be specified.
The parameters are:
1. for
s
you have to enter (
v
,
v
)
(0,00), (10,00)...(80,00), (85,00)
(0,30), (10,30)...(80,30), (85,30)
...
(0,360), (10,360)...(80,360), (85,360)
2. for
s
=
v
you have to enter (
v
,
v
)
(0,00), (10,00)...(80,00), (85,00)
(0,30), (10,30)...(80,30), (85,30)
...
(0,360), (10,360)...(80,360), (85,360)
3. the spherical albedo of the surface
148
SUBROUTINE HAPKBRDF
Function: To generate a BRDF following the Hapke's (1981) model.
Description (from Pinty and Verstraete, 1991): From the fundamental principles of
radiative transfer theory, Hapke (1981) derived an analytical equation for the bidirectional
reflectance function of a medium composed of dimensionless particles. The singly scattered
radiance is derived exactly, whereas the multiply scattered radiance is evaluated from a two-stream
approximation, assuming that the scatterers making up the surface are isotropic. The bidirectional
reflectance of a surface illuminated by the sun from a direction (
s
,
s
), observed from a direction
(
v
,
v
), and normalized with respect to the reflectance of a perfectly reflecting Lambertian surface
under the same condition is given by
(
s
,
s
,
v
,
v
)=
4
1
s
+
v
{[1+B(g)]P(g)+H(
s
)H(
v
)-1}
where
s
=cos(
s
) and
v
=cos(
v
)
g, the phase angle between the incoming and the outgoing rays is defined as:
cos(g)=cos(
s
)cos(
v
)+sin(
s
)sin(
v
)cos(
s
-
v
)
B(g) is a backscattering function that accounts for the hot spot effect
B(g)=
S
H
(0)
P(0)
1
[1+(1/h)tan(g/2)]

H(x) is a function to account for multiple scattering
H(x)=
1+2x
1+2(1-w)
1/2
x

In these equations:
149
is the average single scattering albedo of the particles,
P(g) is the average of the phase function of the particles, it is computed here by the Heyney
and Greenstein's function:
P(g)=
1-
2
1+
2
+2cos(g)

with is the asymmetry factor ranging from -1 (backward scattering) to +1.
S(0) is the amplitude of the hot spot,
h is the width of the hot spot.
The parameters are:
1- , , S(0), h,
where is the average single-scattering albedo of the scatterers, the asymmetry factor for the
phase function, S(0) amplitude of the hot spot, and h width of the hot spot
0.00
0.05
0.10
0.15
0.20
0.25
-80.0 -60.0 -40.0 -20.0 0.0 20.0 40.0 60.0 80.0
C
a
l
c
u
l
a
t
e
d

R
e
f
l
a
c
t
a
n
c
e
s
Viewing angle in the principal plane
backward forward
sun at 60
sun at 00
Clover Patch
=0.101
=-0.263
h=0.046
S(0)=0.589
References
B.W. HAPKE, Bidirectional reflectance spectroscopy - 1. Theory, J. Geophys. Res. , 86, 3039-3054,
1981.
B.W. HAPKE, Bidirectional reflectance spectroscopy - 4. The extinction coefficient and the
opposition effect, Icarus, 67, 264-280, 1986.
B. PINTY, and M. VERSTRAETE, Extracting Information on Surface Properties From
Bidirectional Reflectance Measurements, J. Geophys. Res., 96, 2865-2874, 1991.
150
SUBROUTINE IAPIBRDF
Function: To generate a BRDF following the Iaquinta and Pinty's (1994) model.
Description (from Iaquinta and Pinty, 1994): The model presents an improvement of the
original model by Verstraete et al., 1990, in order to account for the effects due to an underlying
soil below the vegetation canopy. The singly scattered component is solved exactly using an
analytical hot-spot description. The multiply scattered component is approximated on the basis of
Discrete Ordinates Method reduced to a one-angle problem.
The reflectance field is split into three main components which are unscattered (0), singly
scattered (1) and multiply scattered (M) by the leaves:
tot
(
o
,)=
0
(
o
,)+
1
(
o
,)+
M
(
o
)
where the canopy is illuminated from the direction
o
(
o
=cos(
o
),
o
) by the direct solar radiation,
and observed from the direction (=cos(),).
The single scattering by canopy elements
1
(
o
,) is given by
1
(
o
,)=
(
o
)
| |
o
0
LT
T
o
(L)T(L)dL
where (
o
) is the area scattering phase function (bi-Lambertian)
L is the leaf area index (0<L<L
T
)
T
o
(L) is the transmission of direct solar radiation through the canopy layers above the
level L, it can be written as follow
T
o
(L)=exp(-
G(
o
)
| |
o
)
T(L) is the transmission of scattered radiation, it can be written as follow
T(L)=exp(-
G(
o
)
V
2
(
o
,,L)
V(,L)
L)
151
where
V
2
(
o
,,L)
V(,L)
(1-
4
3
)
L
L
i
ifL<L
i
=1-
4
3
L
i
L
ifLL
i
with
L
i
=
2r
tan
2
(
o
)+tan
2
()-2tan(
o
)tan()cos(
o
-)

denotes the leaf area density [m
2
m
-3
] and r [m] is the radius of the
sun-flecks on the illuminated leaf
The second component of the reflectance is the uncollided radiation
0
(
o
,) (first order
reflectance from the soil), which is written
0
(
o
,)=R
s
T
o
(L
T
)T(L
T
)=R
s
exp{-(
G(
o
)
| |
o
+
G()
V
2
(
o
,,L
T
)
V(,L
T
)
)L
T
}
where R
s
is the soil albedo
Using a canopy transport equation reduced to a one-angle problem and assuming isotropic
scattering, the multiply scattered radiation exiting at the top of the canopy is given by
M
(
o
)=
1
| |
o
0
1
I
M
(0,')'d'
where I
M
is the multiply scattered intensity which means photons which have been scattered two or
more times in the finite canopy.
In the above equations, the extinction coefficient of the radiation transport equation uses the G-
function. Physically G(
p
) is the leaf area projected to the direction
p
by a unit leaf area in the
canopy
G(
p
)=
1
2
2+
g
L
(
L
)| |
L
.
p
d
L
where g
L
(
L
) is the probability density of the distribution of leaf normals with respect to the
upward hemisphere (its computation is depending of the input parameter ild)
152
The area scattering phase function (
o
) is given by
1
(
'
)=
1
2
2+
g
L
(
L
)| |
'
.
L
f(
'
,
L
)d
L
where f(
'
,
L
) is the leaf scattering distribution function. Here, it is assuming that the
leaves follow a bi-Lambertian scattering model and then f(
'
,
L
) can be written
f(
'
,
L
)=
{
r
L| | .
L
if(.
L
)('.
L
)<0
t
L| | .
L
if(.
L
)('.
L
)>0

with r
L
the leaf reflection coefficient and t
L
the leaf transmission coefficient
The parameters are :
1- ild, ihs
2- L
t
, 2r
3- r
L
, t
L
, R
s
where ild is the leaf angle distribution (1=planophile, 2=erectophile, 3=plagiophile,
4=extremophile, 5=uniform), ihs a hot spot describter (0=no hot-spot, 1=hot spot), L
t
the leaf area
index in [1. - 15.], 2r a hot-spot parameter in [0. (no hot-spot) - 2.0], r
L
the leaf reflection
coefficient in [0., 0.99], t
L
the leaf transmission coefficient in [0., 0.99], R
s
the soil albedo in [0.,
0.99]
153
0.015
0.020
0.025
0.030
0.035
0.040
0.045
-80.0 -60.0 -40.0 -20.0 0.0 20.0 40.0 60.0 80.0
C
a
l
c
u
l
a
t
e
d

R
e
f
l
e
c
t
a
n
c
e
Viewing angle in principal plane
backward forward
s
=30
L
T
=3.
R
s
=0.2
r
L
=0.0607 & t
L
=0.0429
2r=0.2
Erectophile
no hot-spot
hot-spot
References
J. IAQUINTA, and B. PINTY, Adaptation of a bidirectional reflectance model including the hot-
spot to an optically thin canopy, Proceedings of the VI
e
International Colloquium: Physical
measurements and signatures in remote sensing. Val d'Isre, France, January 17-21.
M. VERSTRAETE, B. PINTY, and R.E. DICKINSON, A Physical Model of the Bidirectional
Reflectance of Vegetation Canopies- 1. Theory, J. Geophys. Res., 95, 11755-11765, 1990.
154
SUBROUTINE MINNBRDF
Function: To generate a BRDF following the Minnaert's (1941) model.
Description: Using the optical reciprocity principle, Minnaert (1941) defined a simple model
to compute the BRDF. The reflectance is written
(
s
,
v
,)=
L
k+1
2
(cos(
v
)cos(
s
))

k-1
where
L
is the albedo of the surface and k the surface parameter. For k=1, the BRDF
corresponds to an ideal Lambertian surface; k=0 corresponds to a minimum reflectance at nadir and
k=2 at a maximum reflectance at nadir.
The parameters are:
1- k,
L
0.00
0.05
0.10
0.15
0.20
-80.0 -60.0 -40.0 -20.0 0.0 20.0 40.0 60.0 80.0
k=2
k=1
k=0
C
a
l
c
u
l
a
t
e
d

R
e
f
l
e
c
t
a
n
c
e
Viewing angles in principal plane
L
=0.10
s
=45
References
M. MINNAERT, The reciprocity principle in lunar photometry, Astrophy. J., 93 , 403-410, 1941
155
SUBROUTINE OCEABRDF (and OCEATOOLS)
Function: To compute the directional reflectance of the ocean surface according to the
whitecaps, sun glint and pigment concentration influences.
Description: In the solar spectral range, the reflectance of the ocean surface
os
() can be
assumed for a given set of geometrical condition
s
,
v
and as the sum of three components
dependent of the wavelength (Koepke, 1984):
os
(
s
,
v
, , )
wc
() +{1 W}.
gl
(
s
,
v
, , ) +{1
wc
()}.
sw
(
s
,
v
, , )
where
wc
() is the reflectance due to the whitecaps
gl
() is the specular reflectance at the ocean surface
sw
() is the scattered reflectance emerging from the sea water
W is the relative area covered with whitecaps, for water temperature greater than
14C it can be expressed from the wind speed ws (Monahan and O'Muircheartaigh,
1980) by W=2.9510
-6
.ws
3.52
1-Reflectance of whitecaps
wc
( )
According to Koepke, 1984, "the optical influence of the whitecaps is given by the product
of the area of each individual whitecaps W with its corresponding reflectance
f
(). However, the
area of an individual whitecap increases with its age while its reflectance decreases. Since
whitecaps of different ages are taken into consideration in the W values, the combination of W with
f
() gives
wc
() values that are too high". Thus Koepke defines, in place of
f
(), an effective
reflectance of ocean foam, patches
ef
() nearly independent of wind speed by the relation
wc
() W.
ef
() W.f
ef
.
f
()
where f
ef
is the efficiency factor slightly dependent of the wind speed but independent of the
wavelength (f
ef
=0.4. 0.2).
The figure 1 shows the reflectance of the whitecaps as a function of the wind speed in the
visible spectral range; in this range the effective reflectance
ef
() is a constant value (22 11)%.
2-Reflectance of the sun glint
gl
( ) (SUBROUTINE SUNGLINT)
156
Cox and Munk (1954,1955) made measurements of the sun glitter from aerial photographs.
They defined a many-faceted model surface whose wave-slopes vary according to isotropic and
anisotropic Gaussian distribution with respect of the surface wind.
Let us consider the system of coordinates (P,X,Y,Z) where P is the observed point, Z the
altitude, PY is pointed to the sun direction and PX to the direction perpendicular to the sun plane. In
this system the surface slope is defined by its two components Z
x
and Z
y
by
x Z
Z
X
sin()tan()
y Z
Z
Y
cos()tan()
where is the azimuth of the ascent (clockwise from the sun) and the tilt. Using spherical
trigonometry Z
x
and Z
y
can be related to the incident and reflected directions (/2
s
and
v
0)
through
x Z
sin(
v
)sin(
s

v
)
cos(
s ) + cos(
v )
y Z
sin(
s
) + sin(
v
) cos(
s

v
)
cos(
s ) + cos(
v )
In the case of an anisotropic distribution of slope components (dependent of the wind
direction), let us consider new principal axes (P;X',Y',Z'=Z) defined by a rotation of from the sun
(P;X,Y,Z) system with PY' parallel to the wind direction (related clockwise from the North by
w
,
then
s

w
). The slope components are now expressed as
Z
x
'=cos().Z
x
+sin().Z
y
Z
y
'=-sin().Z
x
+cos().Z
y
and the slope distribution is expressed by a Gram-Charlier series as
P(
x
'
Z ,
y
'
Z )
1
2
x
'
y
'
exp(
2
+
2
2
) 1
1
2
{
C
21
(
2
1)
1
6
C
03
(
3
) +
+
1
24
C
40
(
4
6
2
+ 3) +
1
4
C
22
(
2
1)(
2
1)+
1
24
C
04
(
4
6
2
+ 3)}
where =Z
x
'/
x
' and =Z
y
'/
y
'

x
' and
y
' are the rms values of Z
x
' and Z
y
', the skewness coefficients C21 and
C03, and the peakedness coefficients C40, C22 and C04 has been defined by Cox
and Munck for a clean (uncontaminated) surface as follows
157
x
' 2
0.003 + 0.00192ws t 0.002
y
' 2
0.00316ws t 0.004
21
C 0.01 0.0086ws t 0.03
03
C 0.04 0.033ws t 0.12
40
C 0.40 t 0.23
22
C 0.12 t 0.06
04
C 0.23 t 0.41
And the directional reflectance is written
gl
(
s
,
v
,
s
,
v
)
P(Z
x
'
,Z
y
'
) R(n,
s
,
v
,
s
,
v
)
4 cos(
s )cos(
v ) cos
4
()
where R(n,
s
,
v
,
s
,
v
) , defined below, is the Fresnel's reflection coefficient (n is the complex
refractive index of the sea water, see below).
Figure 2 shows computations of
gl
for a wind speed of 05 and 15m/s, equal to 00, 90,
180, and 270, a solar zenith angle of 30 and a wavelength of 0.550m.
3-Fresnel's reflection coefficient (SUBROUTINE FRESNEL and INDWAT)
The coefficient R(n, s , v,
s
,
v
) is computed (Born and Wolf, 1975) involving the
absorption of the water (n n
r
in
i
) as
R(n,
s
,
v
,
s
,
v
)
1
2
(n
r
2
n
i
2
)cos(
i
) u
[ ]
2
+ 2n
r
n
i
cos(
i
) + v
[ ]
2
(n
r
2
n
i
2
)cos(
i
) + u
[ ]
2
+ 2n
r
n
i
cos(
i
) v [ ]
2
with
u
2
1
2
{n
r
2
n
i
2
sin
2
(
i
) + n
r
2
n
i
2
sin
2
(
i
)
[ ]
2
+ 4n
r
2
n
i
2
}
v
2
1
2
{ n
r
2
n
i
2
sin
2
(
i
) + n
r
2
n
i
2
sin
2
(
i
)
[ ]
2
+ 4n
r
2
n
i
2
}
cos(
i
)
1
2
[1+ cos(
s
)cos(
v
) + sin(
s
)sin(
v
)sin(
s
v
)]
sin(
i
)
1
2
[1 cos(
s
)cos(
v
) + sin(
s
)sin(
v
)sin(
s
v
)]
In 6S, the complex index of refraction of the sea water (we assume that the outside medium
is vacuum) is deduced from the complex index of refraction of the pure water given by Hale and
Querry, 1973. By default, we assume a typical sea water (Salinity=34.3ppt, Chlorinity=19ppt) as
reported by Sverdrup (1942, p. 173). McLellan (1965, p. 129) reported on measurements of the
index of refraction an increase as a function of the chlorinity. For a chlorinity of 19.0 ppt, the
increase was found to have a value of +0.006 and to be linear with the salt concentration C
sal
(see
158
also Friedman, 1969). For the extinction coefficient, Friedman, 1969, reported that no correction is
required between 1.5 and 9 m.
Then in 6S we apply a correction n
r
of +0.006 on the index of refraction of pure water and no
correction n
i
for the extinction coefficient. Therefore the user is able to enter his own salt
concentration. In that case, a linear interpolation is assumed to correct the index of refraction of the
pure water, so that n
r
=0 for C
sal
=0ppt and n
r
=+0.006 for C
sal
=34.3ppt.
4-Reflectance emerging from the sea water
sw
( ) (SUBROUTINE MORCASIWAT for R
w
)
The reflectance emerging from sea water (also called remote sensing reflectance of the sea
water)
sw
(
s
,
v
,,) is the reflectance as observed just above the sea surface (level 0
+
). This
reflectance can be related to the reflectance R
w
which is the ratio of the upwelling to downwelling
radiance just below the sea surface (level 0
-
). If we assume the ocean as a Lambertian reflector
sw
(
s
,
v
,,) can be expressed by the relation:
sw
(
s
,
v
, , )
1
n
2
R
w
().t
d
(
s
).t
u
(
v
)
1 a.R
w
()
where:
t
d
is the transmittance for the downwelling radiance, and is expressed to the Fresnel reflectance
coefficient R
a-w
(
s
,
d
,) for the air-water interface by the relation:
t
d
(
s
) 1 R
aw
(
s
,
d
a
, ).cos(
d
a
).sin(
d
a
).d
d
a
.d
0
/2
0
2
The angle
d
a
represents (see Figure 3) the zenithal angle of the reflected solar beam according to
the wave-slopes distribution (Cox and Munk's model, see below).
t
u
is the transmittance for the upwelling radiance, and is expressed to the Fresnel reflectance
coefficient R
w-a
(
v
,
u
,) for the water-air interface by
t
u
(
v
) 1 R
wa
(
v
,
u
w
, ).cos(
u
w
).sin(
u
w
).d
u
w
.d
0
/2
0
2
The angle
u
w
represents (see Figure 3) the zenith angle in the water of the upwelling beam
according to the Fresnel and Snell's law: n
air
sin(
air
)=n
sea
sin(
sea
) and to the wave-slopes
distribution.
a is defined by
a 1 t
u
(
v
).cos(
v
).sin(
v
).d
0
/2
v
In order to minimize computations we adopted a constant value of a=0.485. In theory, the value of a
depends on wind speed and water index of refraction. In practice, this value varies very little with
159
wind speed (see table 2 of Austin, 1974) and the index of refraction of water in that range of
wavelength (0.4m-0.7m) is almost constant and taken equal to 1.341.
As described above the irradiance reflectance R
w
() is the ratio of the upwelling spectral
irradiance E
u
() to the downwelling irradiance E
d
() just below the surface. This ratio is
particularly dependent on the inherent optical properties of the sea water: the total absorption
coefficients a() [m
-1
] and the total backscattering coefficient b
b
() [m
-1
]. For example, Morel and
Prieur, 1977, have shown within a good approximation (when a() <<1) that it can be expressed as:
R
w
() 0.33
b
b
()
a()
According to Morel, 1988, "in many situations phytoplankton and their derivative, and
detrital products (mainly particulate, but also dissolved) play a predominant role in determining the
optical properties of oceanic waters. These waters are classified (by Morel) as "case I" waters and
are opposed to "case II" waters for which sediments, or dissolved yellow substance, make an
important or dominant contribution to the optical properties". Here we use the so called "case I
waters" (defined in a range from 0.4 to 0.7 m) which roughly corresponds to the case I, case IA,
case IB, case II, and case III of the Jerlov's chart of optical water type (Jerlov, 1951, 1976). For the
so called "Case I waters", Morel splits the total backscattering coefficient into
b
b
()
1
2
b
w
() +

b
b
().b
where:
b
w
() is the molecular scattering coefficient of water and is given Figure 4

b
b
() is the ratio backscattering/scattering coefficient of the pigments and is related to the
pigment concentration C (Chl a + Pheo a, in mg.m
-3
) and the wavelength (in m) by

b
b
() 0.002 + 0.02(0.5 0.25 log
10
C)
0.550
b is the scattering coefficient of pigment expressed by

b=0.3C
0.62
Also according to Morel's "Case I waters", the total absorption coefficient is written as
a()=u().K
d
()
where:
u() is computed as follow
u() 0.90
1 R
w
()
1+ 2.25 R
w
()
K
d
() is the total diffuse attenuation coefficient for downwelling irradiance and is given by
K
d
()=K
w
()+
c()C
e()
160
with K
w
() (the diffuse attenuation coefficient for pure oceanic water),
c() and e() are tabular

values. We report Figure 5 the computations of K
d
() for several pigment concentrations.
Always according to the Morel's model (Case I waters) the computation of the reflectance
R
w
() is only dependent of the pigment concentration C. The Figure 6 shows the computed
reflectance R
w
in the range from 0.4 to 0.7 m for different concentrations C.
If the only information you have is the water type following the Jerlov's chart, you can use
the approximate C values given by Morel, 1988:
0-0.01 0.05 0.1 0.5 1.5-2 mg.m
-3
I IA IB II III
The Parameters are:
1- ws,
w
, C
sal
, C
with: ws is the wind speed (in m/s)
w
is the direction of the wind (clockwise from the North)
C
sal
is the salt concentration (in ppt). If C
sal
< 0 then C
sal
=34.3ppt by default
C is the pigment concentration (Chl a + Pheo a in mg.m
-3
)
161
0.000
0.005
0.010
0.015
0.020
0 5 10 15 20
W
h
i
t
c
a
p
s

R
e
f
l
e
c
t
a
n
c
e
Wind Speed
Figure 1: Whitecaps Reflectance as defined by Koepke, 1984,
in the visible spectral range.
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
-60 -40 -20 0 20 40 60
=00
=90-270
=180
g
l
ws = 05 m/s
ws = 15 m/s
Figure 2: Reflectance of the sunglint in the principal plane for a
wind speed of 5 and 15 m/s and for several =
s
-
w
. The solar zenith
angle has been up to 30 and the wavelength to 0.550m.
162
d1
d3
d2
Sea
Surface
Level 0
+
Level 0
-
Sea
Water
R
w
v
a
a
a
u3
w
u2
w
u1
w
E
d
E
u
Figure 3
0.000
0.002
0.004
0.006
0.008
0.010
0.4 0.45 0.5 0.55 0.6 0.65 0.7
s
p
e
c
t
r
a
l

m
o
l
e
c
u
l
a
r

s
c
a
t
t
e
r
i
n
g

c
o
e
f
f
i
c
i
e
n
t

b
w
(
)
( m)
Figure 4: spectral molecular scattering coefficients (m
-1
)
163
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70
t
o
t
a
l

d
i
f
f
u
s
e

a
t
t
e
n
u
a
t
i
o
n

c
o
e
f
f
i
c
i
e
n
t

(m)
10.0
3.0
1.0
0.0
Figure 5: Diffuse attenuation coefficients (m
-1
) as modeled in Morel's case I waters
for several pigment concentration (0.0-0.03-0.1-0.3-1-3-10 mg.m
-3
)
10
- 4
10
- 3
10
- 2
10
- 1
350 400 450 500 550 600 650 700 750
R
w
(
)
(nm)
0.3
3.0
1.0
0.1
0.03
0.0
10.0
Figure 6: Spectral irradiance reflectances as defined by the Morel's model related to the "Case I
waters" for several pigment concentration (0.0-0.03-0.1-0.3-1-3-10 mg.m
-3
)
164
References
R.W. AUSTIN, and T.J. PETZOLD, Spectral dependence of the diffuse attenuation coefficient of
light in ocean water, Opt. Eng., 25, 471-479, 1986.
M. BORN, and E. WOLF, Principles of Optics - fifth edition, Pergamon Press, New-York, 1975
C. COX, and W. MUNK, Statistics of the sea surface derived from sun glitter, J. Marine Res., 13,
198-227, 1954.
C. COX, and W. MUNK, Measurement of the roughness of the sea surface from photographs of the
sun's glitter, J. Opt. Soc. Am., 44, 838-850, 1954.
C. COX, and W. MUNK, Some problems in optical oceanography, J. Marine Res., 14, 63-78, 1955.
D. FRIEDMAN, Infrared Characteristics of Ocean Water (1.5-15), Appl. Opt. , 8-10, 2073-2078,
1969.
G.M. HALE, and M.R. QUERRY, Optical Constants of Water in the 200 nm to 200 m
Wavelength Region, Appl. Opt., 12-3, 555-563, 1973.
N.G. JERLOV, Optical Studies of Ocean Water, Rep. Swed. Deep. Sea. Exped., 1947 1948, 3, 1-19,
1951.
N.G. JERLOV, Marine Optics, Elsevier Oceanogr. Ser., vol. 14, 231 pp., Elsevier, Amsterdam,
1976.
P. KOEPKE, Effective Reflectance of Oceanic Whitecaps, Appl. Opt., 23-11, 1816-1824, 1984.
H.J. McLELLAN, Elements of physical Oceanography, Pergamon Press, Inc., New-York, 1965.
E.C. MONAHAN, and I. O'MUIRCHEARTAIGH, Optimal Power-Law Description of Oceanic
Whitecap Dependence on Wind Speed, J. Phys. Ocean, 10, 2094, 1980.
A. MOREL, Optical Modeling of the Upper Ocean in Relation to its Biogenous Matter Content
(Case I Waters), J. Geophys. Res., 93-C9, 10749-10768, 1988.
A. MOREL, and L. PRIEUR, Analysis of variations in ocean color, Limnol. Oceanogr., 22, 709-
722, 1977.
H. V. SVERDRUP, M.W. JOHNSON, and R.H. FLEMING, The Ocean, Prentice-Hall, Inc.,
Englewood Cliffs, N.J., 1942.
165
SUBROUTINE RAHMBRDF
Function: To generate a BRDF following the Rahman et al.'s model.
Description (from Raman et al., 1993) : The model follows a semi-empirical approach and is
designed to be applicable to arbitrary natural-surfaces both in visible and near-infrared using 3
parameters.
The bidirectional reflectance (
s
,
v
,
s
-
v
) is assumed to be:
(
s
,
v
,
s
,
v
)
0
(cos
s
)
k1
(cos
v
)
k1
(cos
s
+ cos
v
)
1k
F(g) 1 + R(G) [ ]
where: -
0
is, according to the authors, an arbitrary parameter characterizing the intensity of the
reflectance of the surface cover, but it should not be taken as a single scattering albedo
or as a normalized reflectance: the only constraint on it is
0
0,
-k indicates the level of anisotropy of this surface,
-F(g) is a function to moderate the overall contributions in the forward and backward
scattering (modified Heyney and Greenstein's function):
F(g)
1
2
[1+
2
2cos( g)]
1.5
with g the phase angle defined by:
cos(g) cos(
s
)cos(
v
) + sin(
s
)sin(
v
)cos(
s
v
)
and an asymmetry factor controling the relative amount of forward
(0+1) and backward scattering (-10)
-R(G) is a function to account for the hot spot
R(G)
1
0
1+ G
with G a geometrical factor expressed by:
G (tan
s
)
2
+ (tan
v
)
2
2tan
s
tan
v
cos(
s
v
)
166
Parameters are:
1-
0
, , k
with
0
=intnsity of the reflectance of the surface cover
=asymmetry factor used in the modified Heyney and Greenstrein's function
k=structural parameter
0.00
0.02
0.04
0.06
0.08
0.10
0.12
-80.0 -60.0 -40.0 -20.0 0.0 20.0 40.0 60.0 80.0
C
a
l
c
u
l
a
t
e
d

R
e
f
l
e
c
t
a
n
c
e
Sun at 60
Sun at 00
0
=0.012
=-0.391
k=0.811
Clover Patch
Backward Forward
References
H. RAHMAN, M. VERSTRAETE, and B. PINTY, A coupled Surface-Atmosphere Reflectance
(CSAR) Model. Part1: Model description and inversion on synthetic data, Submitted at J.
Geophys. Res., 1993
H. RAHMAN, B. PINTY, and M. VERSTRAETE, A coupled Surface-Atmosphere Reflectance
(CSAR) Model. Part2: A Semi-empirical Surface Model usable with NOAA/AVHRR Data,
Submitted at J. Geophys. Res., 1993
167
SUBROUTINE ROUJBRDF
Function: To generate a BRDF following the Roujean et al.'s (1992) model.
Description (from Roujean et al., 1992): The model follows a semi-empirical approach and
is designed to be applicable to heterogeneous surfaces. It considers that the observed surface
bidirectional reflectance is the sum of two main processes operating at a local scale:
- A diffuse reflection component taking into account the geometrical structure of opaque
reflectors on the surface, and shadowing effects
- A volume scattering contribution by a collection of dispersed facets which simulates the
volume scattering properties of canopies and bare soils.
This combinaison is made by assuming that the bidirectional reflectance (
s
,
v
,) can be
expressed as:
(
s
,
v
,)=
geom
+(1-)
vol
where is an empirical coefficient which characterizes the relative weight of the geometric and
volume component in the final bidirectional signature.
Then the bidirectional reflectance model may be written
(
s
,
v
,)=k
0
+k
1
f
1
(
s
,
v
,
r
)+k
2
f
2
(
s
,
v
,
r
)
where: = | |
s
-
v

k
0
, k
1
and k
2
are related to basic macroscopic properties of the surface
k
0
=
0
[+(1-)e
-bF
]+
r
3
(1-e
-bF
)(1-)
k
1
=
hl
S

0
k
2
=r(1-e
-bF
)(1-)
with
0
, the background and protrusion reflectance,
h, the average height of the surface protrusions,
l, the average length of the surface protrusions,
S, the horizontal surface associated with each protrusion,
r, the facet reflectance,
F, the facet area index.
f1 and f
2
are simple analytic functions of the solar and viewing angles
168
f
1
(
s
,
v
,)=
1
2
{(-)cos()+sin()}tg(
s
)tg(
v
)
-
1
{tg(
s
)+tg(
v
)+ tg
2
(
s
)+tg
2
(
v
)-2tg(
s
)tg(
v
)cos() }
f
2
(
s
,
v
,)=
4
3
1
cos(
s
)+cos(
v
)
{(
2
-)cos()+sin()}-
1
3

with , the phase angle defined by
cos() = cos(
s
) cos(
v
) + sin(
s
) sin(
v
) cos()
The parameters are:
1- k
0
, k
1
, k
2
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
-80.0 -60.0 -40.0 -20.0 0.0 20.0 40.0 60.0 80.0
C
a
l
c
u
l
a
t
e
d

R
e
f
l
e
c
t
a
n
c
e
s
backward
forward
sun at 45
sun at 30
Plowed field
k
0
= 0.243
k
1
= 0.073
k
2
= 0.642
References
J.L. ROUJEAN, M. LEROY, and P.Y. DESCHAMPS, A bidirectional reflectance model of the
Earth surface for the correction of remote sensing data, J. Geophys. Res. , 97, 20455-20468,
1992.
169
SUBROUTINE VERSBRDF
Function: To generate a BRDF following the Verstraete et al.'s (1990) model.
Description (from Pinty and Verstraete, 1991): Using the basic framework previously
suggested by Hapke (1981) (see HAPKBRDF), Verstraete et al. (1990) developed a model for
predicting the bidirectional reflectance exiting from a simple vegetation canopy. They concentrate
on the case of a fully covered, homogeneous and semi-infinite canopy made of leaves only. With
the same notations as for Hapke's model, the parametric version of the derived model (see Pinty et
al., 1990) is as follow:
(
s
,
s
,
v
,
v
)=
s
+
s
v
{[1+P
v
(G)]P(g)+H(
s
/
s
)H(
v
/
v
)-1}
where
is average single-scattering albedo of the particle making up the surface

s
=cos(
s
) and
v
=cos(
v
)

s
and
v
describe the leaf orientation distribution for the illumination and viewing angles,
respectively. There are 3 possible options (see below, line1-opt3)
-
s
and
v
are entered by the user
- is obtained from Goudriaan's (1977) parameterization
x
=
1
+
2

x
with
1
= 0.5 - 0.6333
l
- 0.33
l
2
and
2
= 0.877 (1 - 2
1
)
- is obtained from the Dickinson et al.'s (1990) correction to Goudriaan's (1977)
parameterization
x
=
1
+
2

x
with
1
= 0.5 - 0.489
l
- 0.11
l
2
and
2
= 1 - 2
1
where
l
is a function of the leaf angle distribution in the canopy, and varies from -0.4
for an erectophile canopy to +0.6 for a planophile canopy. The equal probability for all
leaf orientations is given by
l
=0.
Pv(G) is the function that accounts for the joint transmission of the incoming and outgoing
radiation and thereby also for the hot spot phenomenon.
P
v
(
s
,
s
,
v
,
v
)=
1
1+V
p
(
s
,
s
,
v
,
v
)

170
with V
p
(
s
,
s
,
v
,
v
) is a function defined by
4(1-
4
3
)
1
2r
v
tan
2
(
s
)+tan
2
(
v
)-2tan(
s
)tan(
v
)cos(
s
-
v
)
r is the radius of the Sun flecks on the inclined scatterers (in m)
is the scatterer area density of the canopy (in m
2
.m
-3
)
P(g) is the average of the phase function of the particles. There are 3 options to compute
P(g) (see below, line1-opt4):
-the case of an isotropic phase function
P(g) = 1
-the empirical function introduced by Henyey and Greenstein (1941)
P(g)=
1-
2
1+
2
+2cos(g)

-the phase function is approximated by a Legendre polynomial function
P(g)=1+cos(g)+L
2
3cos
2
(g)-1
2

where is the asymmetry factor ranging from -1 (backward scattering) to +1, g the
phase angle between the incoming and the outgoing rays defined as
cos(g) = cos(
s
) cos(
v
) + sin(
s
) sin(
v
) cos(
s
-
v
),
and L
2
the second coefficients of the Legendre polynomial.
H(x) is a function to account for multiple scattering (see below, line1-opt5)
- for single scattering
H(x)=0
- for multiple scattering
H(x)=
1+2x
1+2(1-)
1/2
x

171
There are three lines of input parameters:
line 1- choice of options : opt3, opt4, opt5
line 2- structural parameters : str1, str2, str3, str4
line 3- optical parameters : optics1, optics2, optics3
line 1 :
opt 3- 0 for given values of
1 for Goudriaan's parameterization of
2 for Dickinson et al.'s correction to Goudriaan's parameterization of
opt 4- 0 for isotropic phase function
1 for Heyney and Greenstein's phase function
2 for Legendre polynomial phase function
opt 5- 0 for single scattering only
1 for Dickinson et al.'s parameterization of multiple scattering
line 2 :
str1- leaf area density (in m
2
m
-3
)
str2- radius of the sun flecks on the scatterer (in m)
str3- leaf orientation parameter
if opt3=0 then str3=
s
if opt3=1 or 2 then str3=
l
str4- leaf orientation parameter (continued)
if opt3=0 then str4=
v
if opt3=1 or 2 then str4 is not used
line 3 :
optics 1- single scattering albedo, value between 0.0 and 1.0
optics 2- phase function parameter
if opt4=0 then this input is not used
if opt4=1 then asymmetry factor, value between -1.0 and 1.0
if opt4=2 then first coefficient of Legendre polynomial
optics 3- second coefficient of Legendre polynomial (if opt4=2)
172
0.00
0.05
0.10
0.15
0.20
-80.0 -60.0 -40.0 -20.0 0.0 20.0 40.0 60.0 80.0
C
a
l
c
u
l
a
t
e
d

R
e
f
l
e
c
t
a
n
c
e
Viewing Angle in the principal plane
Backward forward
Sun at 00
Sun at 60
Clover Patch
=0.099
=-0.391
l
=0.115
2r=0.277
References
R.E. DICKINSON, B. PINTY, and M. VERSTRAETE, Relating surface albedos in gcm to
remotely sensed data, agricultural and forest meteorology, 52, 109-131, 1990.
J. GOUDRIAAN, Crop micrometeorology: a simulation study (Wageningen: Wageningen Centre
for Agricultural Publishing and Documentation), 1977.
L. G. HENYEY AND J. L. GREENSTEIN,, Diffuse Radiation in the Galaxy. Astrophysic Journal,
93, 70, 1941.
B. PINTY, M. VERSTRAETE, and R.E. DICKINSON, A Physical Model of the Bidirectional
Reflectance of Vegetation Canopies- 1. Inversion and Validation, J. Geophys. Res., 95,
11767-11775, 1990.
B. PINTY, and M. VERSTRAETE, Extracting Information on Surface Properties From
Bidirectional Reflectance Measurements, J. Geophys. Res., 96, 2865-2874, 1991.
M. VERSTRAETE, B. PINTY, and R.E. DICKINSON, A Physical Model of the Bidirectional
Reflectance of Vegetation Canopies- 1. Theory, J. Geophys. Res., 95, 11755-11765, 1990.
173
SUBROUTINE WALTBRDF
Function: To generate a BRDF following the Walthall et al.'s (1985) model.
Description (from Walthall et al., 1985): Using a deterministic model, Walthall et al. (1985)
have simulated different canopy reflectance distributions. The 2-D contours of this distribution
appearsto be similar to the shape of the limacon of Pascal. Using the simple limacon equation other
equation forms were used in an attempt to fit the 3-D reflectance surface directly. They found
satisfactory results with the following equation:
(
s
,
v
,
s
,
v
) = a
v
2
+ b
v
cos(
v
-
s
) + c
where is the reflectance at a given view zenith
v
and view azimuth
v
look angles; a, b, and c are
coefficients derived using a linear least-squares fitting procedure.
The model has to be slightly modified to match the reciprocity principle. The reflectance is written:
(
s
,
v
,
s
,
v
) = a
s
2
v
2
+ a' (
s
2
+
v
2
) + b
s
v
cos(
v
-
s
) + c
The parameters are:
1- a, a', b, c
0.02
0.03
0.04
0.05
0.06
-80.0 -60.0 -40.0 -20.0 0.0 20.0 40.0 60.0 80.0
backward forward
a = 0.00606
a'= 0.00523
b = 0.01127
c = 0.02425
sun at 17
sun at 40
LAI=4
References
C.L. WALTHALL, J.M. NORMAN, J.M. WELLES, G. CAMPBELL, and B.L. BLAD, Simple
equation to approximate the bidirectional reflectance from vegetative canopies and bare soil
surfaces, Applied Optics, Vol 24, n3, 383-387, 1985.
174
175
DESCRIPTION OF THE SUBROUTINES USED TO UPDATE ATMOSPHERIC
PROFILE (PLANE OR ELEVATED TARGET SIMULATION)
176
177
SUBROUTINE PRESPLANE
Function: Update the atmospheric profile (P(z),T(z),H
2
O(z),O
3
(z)) in case the observer is on
board an aircraft.
Description: Given the altitude or pressure at aircraft level as input, the first task is to
compute the altitude (in case the pressure has been entered) or the pressure (in case the altitude has
been entered) at plane level. Then, a new atmospheric profile is created (P
p
,T
p
,H
2
O
p
,O
3p
) for which
the last level is located at the plane altitude. This profile is used in the gaseous absorption
computation (ABSTRA.f) for the path from target to sensor (upward transmission). The ozone and
water vapor integrated content of the "plane" atmospheric profile are also an output of this
subroutine. The last output is the proportion of molecules below plane level which is useful in
scattering computations (OS.f,ISO.f).
178
SUBROUTINE PRESSURE
Function: Update the atmospheric profile (P(z),T(z),H
2
O(z),O
3
(z)) in case the target is not at
sea level.
Description: Given the altitude of the target in kilometers as input, we transform the
original atmospheric profile (Pressure, Temperature, Water Vapor, Ozone) so that first level of the
new profile is the one at the target altitude. We also compute the new integrated content in water
vapor and ozone, that are used as outputs or in computations when the user chooses to enter a
specific amount of Ozone and Water Vapor.
179
DESCRIPTION OF THE SUBROUTINES
USED TO READ THE DATA
180
181
SUBROUTINE SOLIRR
Function: To read the solar irradiance (in Wm
-2
) from 250 nm to 4000 nm by steps of 5 nm,
The total solar irradiance is put equal to 1372 Wm
-2
. Between 250 and 4000 nm we have 1358 Wm
-
2
.
0
500
1000
1500
2000
0 0.5 1 1.5 2 2.5 3 3.5 4
S
o
l
a
r

I
r
r
a
d
i
a
n
c
e

[
W
/
m
2
]
Wavelength [ m]
Reference:
H. NECKEL, D. LABS, The solar radiation between 3300 and 12500 Solar Physics 90, p. 205-258,
1984.
182
SUBROUTINE VARSOL
Function: To take into account the variation of the solar constant as a function of the Julian
day.
Description: We apply a simple multiplicative factor D
S
to the solar constant C
S
.D
S
is
written as,
D
S
=
1
(1-ecosM)
2

with
M = 0.9856 (J-4)
180

where e = 0.01673 and J is the julian day.
Reference:
G.W. PALTRIDGE, C.M.R. PLATT, Radiative Processes in Meteorology and Climatology,
Development in Atmospheric Science, 5, Elsevier Scientific Publishing Company, New-York,
N.Y. 10017, 1977.
183
SUBROUTINE AVHRR
Function: To read the two spectral bands (red and near infrared ) of Advanced Very High
Resolution Radiometer (AVHRR) on NOAA 6,7,8,9,10 and 11 (extreme wavelengths and spectral
response of the filter function).
0
20
40
60
80
100
0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1
S
p
e
c
t
r
a
l

R
e
s
p
o
n
s
e

[
%
]
Wavelength [ m]
NOAA6
Channel 1
0
20
40
60
80
100
0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1
S
p
e
c
t
r
a
l

R
e
s
p
o
n
s
e

[
%
]
Wavelength [ m]
NOAA6
Channel 2
0
20
40
60
80
100
0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1
S
p
e
c
t
r
a
l

R
e
s
p
o
n
s
e

[
%
]
Wavelength [ m]
NOAA7
Channel 1
0
20
40
60
80
100
0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1
S
p
e
c
t
r
a
l

R
e
s
p
o
n
s
e

[
%
]
Wavelength [ m]
NOAA7
Channel 2
184
0
20
40
60
80
100
0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1
S
p
e
c
t
r
a
l

R
e
s
p
o
n
s
e

[
%
]
Wavelength [ m]
NOAA8
Channel 1
0
20
40
60
80
100
0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1
S
p
e
c
t
r
a
l

R
e
s
p
o
n
s
e

[
%
]
Wavelength [ m]
NOAA8
Channel 2
0
20
40
60
80
100
0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1
S
p
e
c
t
r
a
l

R
e
s
p
o
n
s
e

[
%
]
Wavelength [ m]
NOAA9
Channel 1
0
20
40
60
80
100
0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1
S
p
e
c
t
r
a
l

R
e
s
p
o
n
s
e

[
%
]
Wavelength [ m]
NOAA9
Channel 2
0
20
40
60
80
100
0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1
S
p
e
c
t
r
a
l

R
e
s
p
o
n
s
e

[
%
]
Wavelength [ m]
NOAA10
Channel 1
0
20
40
60
80
100
0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1
S
p
e
c
t
r
a
l

R
e
s
p
o
n
s
e

[
%
]
Wavelength [ m]
NOAA10
Channel 2
185
0
20
40
60
80
100
0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1
S
p
e
c
t
r
a
l

R
e
s
p
o
n
s
e

[
%
]
Wavelength [ m]
NOAA11
Channel 1
0
20
40
60
80
100
0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1
S
p
e
c
t
r
a
l

R
e
s
p
o
n
s
e

[
%
]
Wavelength [ m]
NOAA11
Channel 2
References:
NOAA POLAR ORBITER DATA USERS GUIDE, (1985) U.S. Dept. of Commerce, NOAA,
National Environment Satellite, National Climatic Data Center, Satellite Data Service Division,
World Weather Building, Room 100, Washington D.C., 202333, U.S.A.
DUE C.T., 1982, Optical-Mechanical Active/passive imaging Systems - Volume II, Report number
153200-2-TIII- ERIM Infrared information and Analysis Center, P.O. BOX 8518, Ann. Arbor.,
MI.98107.
SCHNEIDER S.R. and Mc GINNIS D.F., 1982, The NOAA/AVHRR: A new satellite sensor for
monitoring crop growth, Proc. 8th Inter. Symp.on Machine Processing of Remotely Sensed
data, Purdue University, Indiana, p. 250-281.
186
SUBROUTINE GOES
Function: To read the visible spectral bands of the Visible Infrared Spin-Scan Radiometer
(VISSR) on GOES 5 (East) and GOES 4 (West), (extreme wavelengths and spectral response of the
filter function).
0
20
40
60
80
100
0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1
S
p
e
c
t
r
a
l

R
e
s
p
o
n
s
e

[
%
]
Wavelength [ m]
GOES
EAST
0
20
40
60
80
100
120
0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1
S
p
e
c
t
r
a
l

R
e
s
p
o
n
s
e

[
%
]
Wavelength [ m]
GOES
WEST
References:
STAFF MEMBERS, 1980, Visible infrared Spin Scan Radiometer Atmospheric sounder system
description, Santa Barbara Research center, Goleta, Cali., U.S.A.
CORBELL R.P., C.J. CULLAHAN and W.J. KOTSCH, 1976, The GOES/SMS user's Guide: U.S.
Dept. of Commerce, NOAA, NESS, Washington D.C., U.S.A.
187
SUBROUTINE HRV
Function: To read the four spectral bands of High Resolution Visible (HRV1 and 2) on Spot
1 (extreme wavelengths and spectral response of the filter function).
0
20
40
60
80
100
0.4 0.45 0.5 0.55 0.6 0.65 0.7
S
p
e
c
t
r
a
l

R
e
s
p
o
n
s
e

[
%
]
Wavelength [ m]
HRV1
Channel 1
0
20
40
60
80
100
0.5 0.55 0.6 0.65 0.7 0.75 0.8
S
p
e
c
t
r
a
l

R
e
s
p
o
n
s
e

[
%
]
Wavelength [ m]
HRV1
Channel 2
0
20
40
60
80
100
0.7 0.75 0.8 0.85 0.9 0.95 1
S
p
e
c
t
r
a
l

R
e
s
p
o
n
s
e

[
%
]
Wavelength [ m]
HRV1
Channel 3
0
20
40
60
80
100
0.4 0.45 0.5 0.55 0.6 0.65 0.7 0.75 0.8
S
p
e
c
t
r
a
l

R
e
s
p
o
n
s
e

[
%
]
Wavelength [ m]
HRV1
Panchromatique
188
0
20
40
60
80
100
0.4 0.45 0.5 0.55 0.6 0.65 0.7
S
p
e
c
t
r
a
l

R
e
s
p
o
n
s
e

[
%
]
Wavelength [ m]
HRV2
Channel 1
0
20
40
60
80
100
0.5 0.55 0.6 0.65 0.7 0.75 0.8
S
p
e
c
t
r
a
l

R
e
s
p
o
n
s
e

[
%
]
Wavelength [ m]
HRV2
Channel 2
0
20
40
60
80
100
0.7 0.75 0.8 0.85 0.9 0.95 1
S
p
e
c
t
r
a
l

R
e
s
p
o
n
s
e

[
%
]
Wavelength [ m]
HRV2
Channel 3
0
20
40
60
80
100
0.4 0.45 0.5 0.55 0.6 0.65 0.7 0.75 0.8
S
p
e
c
t
r
a
l

R
e
s
p
o
n
s
e

[
%
]
Wavelength [ m]
HRV2
Panchromatique
Reference:
CHEVREL M., M. COURTOIS, and G. WEILL, 1981, The Spot Satellite Remote Sensing Mission
Photogrammetric Engineering and Remote Sensing, Vol. 47, no 8, p. 1163-1171.
189
SUBROUTINE METEO
Function: To read the visible spectral band of the radiometer on Meteosat 2 (extreme
wavelengths and spectral response of the filter function).
0
20
40
60
80
100
0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1
S
p
e
c
t
r
a
l

r
e
s
p
o
n
s
e

[
%
]
Wavelenght [ m]
METEOSAT 2
Reference:
MORGAN, 1981, Introduction to the Meteosat System, ESOC, Darmstadt, R.F.A.
190
SUBROUTINE MSS
Function: To read the four spectral bands of the Multispectral Scanner System (MSS) on
Landsat 5 (extreme wavelengths and spectral response of the filter function).
0
20
40
60
80
100
0.4 0.45 0.5 0.55 0.6 0.65 0.7
S
p
e
c
t
r
a
l

R
e
s
p
o
n
s
e

[
%
]
Wavelength [ m]
MSS4
0
20
40
60
80
100
0.5 0.55 0.6 0.65 0.7 0.75 0.8
S
p
e
c
t
r
a
l

R
e
s
p
o
n
s
e

[
%
]
Wavelength [ m]
MSS5
0
20
40
60
80
100
0.6 0.65 0.7 0.75 0.8 0.85 0.9
S
p
e
c
t
r
a
l

R
e
s
p
o
n
s
e

[
%
]
Wavelength [ m]
MSS6
0
20
40
60
80
100
0.7 0.8 0.9 1 1.1 1.2
S
p
e
c
t
r
a
l

R
e
s
p
o
n
s
e

[
%
]
Wavelength [ m]
MSS7
References:
LANDSAT DATA USERS HANDBOOK (revised) (1979), U.S. Geol. Survey, EROS Data Center,
Sioux Falls, SD 57198.
LANDSAT DATA USERS NOTES (1982), International Land Satellite programs, ibid.
191
SUBROUTINE TM
Function: To read the six visible spectral bands of Thematic Mapper (TM) on Landsat 5
(extreme wavelengths and spectral response of the filter function).
0
20
40
60
80
100
0.4 0.45 0.5 0.55 0.6 0.65 0.7
S
p
e
c
t
r
a
l

R
e
s
p
o
n
s
e

[
%
]
Wavelength [ m]
TM 1
0
20
40
60
80
100
0.4 0.45 0.5 0.55 0.6 0.65 0.7
S
p
e
c
t
r
a
l

R
e
s
p
o
n
s
e

[
%
]
Wavelength [ m]
TM 2
0
20
40
60
80
100
0.5 0.55 0.6 0.65 0.7 0.75 0.8
S
p
e
c
t
r
a
l

R
e
s
p
o
n
s
e

[
%
]
Wavelength [ m]
TM 3
0
20
40
60
80
100
0.7 0.75 0.8 0.85 0.9 0.95 1
S
p
e
c
t
r
a
l

R
e
s
p
o
n
s
e

[
%
]
Wavelength [ m]
TM 4
0
20
40
60
80
100
1.4 1.5 1.6 1.7 1.8 1.9 2
S
p
e
c
t
r
a
l

R
e
s
p
o
n
s
e

[
%
]
Wavelength [ m]
TM 5
0
20
40
60
80
100
1.9 2 2.1 2.2 2.3 2.4 2.5
S
p
e
c
t
r
a
l

R
e
s
p
o
n
s
e

[
%
]
Wavelength [ m]
TM 7
Reference:
MARKHAM B.L. and J.L. BARKER, 1985, Spectral Characterization of the LANDSAT thematic
Mapper sensors, Int. J. Remote Sensing,
192
SUBROUTINE CLEARW
Function: To read a typical spectral reflectance of clear water (sea) from 250 to 4000 nm by
steps of 5 nm.
0
1
2
3
4
5
6
7
0.4 0.5 0.6 0.7 0.8 0.9 1
Clear water (sea)
Wavelengths (microns)
Reference:
M. VIOLLIER, Tldtection des concentrations de seston et pigments chlorophylliens contenus
dans l'Ocan, Thse de Doctorat d'Etat, no 503, 1980.
193
SUBROUTINE LAKEW
Function: To read a typical spectral reflectance of water (lake) from 250 to 4000 nm by steps
of 5 nm.
0
2
4
6
8
10
0.4 0.5 0.6 0.7 0.8 0.9 1
Clear water (lake)
Wavelength [ m]
Reference:
KONDRATYEV K. Ya (1969), Radiation in the atmosphere, Academic Press, N.Y. 10003, U.S.A.
194
SUBROUTINE SAND
Function: To read a typical spectral reflectance of sand from 250 to 4000 nm by steps of 5
nm.
0
5
10
15
20
25
30
35
40
0.4 0.8 1.2 1.6 2
Dry sand
Wavelength [ m]
Reference:
R. STAETTER, M. SCHROEDER, Spectral characteristics of natural surfaces, Proceeding of ten
Int. Conf. on Earth Obs. from Space, 6-11 March 1978, (ESA-SP, 134).
195
SUBROUTINE VEGETA
Function: To read a typical spectral reflectance of a mean green vegetation surface from 250
to 4000 nm by steps of 5 nm.
0
10
20
30
40
50
60
0.5 1 1.5 2 2.5
Clear water (sea)
Clear water (lake)
Vegetation
Dry sand
R
e
f
l
e
c
t
a
n
c
e

[
%
]
Wavelength [ m]
Reference:
Manual of Remote Sensing, Falls Church, Virginia, American Society of Photogrammetry, 1983.
196
SUBROUTINE DICA1
Function: To read the eight coefficients necessary to compute the CO
2
transmission
according to the Malkmus model (see the text). The frequency interval is 2500-5050 cm
-1
, step of
10 cm
-1
.
SUBROUTINE DICA2
Function: Same as DICA1 but for frequency interval 5060-7610 cm
-1
.
SUBROUTINE DICA3
Function: Same as DICA1 but for frequency interval 7620-10170 cm
-1
.
197
SUBROUTINE METH1
Function: To read the eight coefficients necessary to compute the methane transmission
-1
, step of
10 cm
-1
.
SUBROUTINE METH2
Function: Same as METH1 but for frequency interval 5060-7610 cm
-1
.
SUBROUTINE METH3
-1
.
SUBROUTINE METH4
-1
.
SUBROUTINE METH5
-1
.
SUBROUTINE METH6
-1
.
198
SUBROUTINE MOCA1
Function: To read the eight coefficients necessary to compute the CO transmission according
to the Malkmus model (see the text). The frequency interval is 2500-5050 cm
-1
, step of 10 cm
-1
.
SUBROUTINE MOCA2
Function: Same as MOCA1 but for frequency interval 5060-7610 cm
-1
.
SUBROUTINE MOCA3
-1
.
SUBROUTINE MOCA4
-1
.
SUBROUTINE MOCA5
-1
.
SUBROUTINE MOCA6
-1
.
199
SUBROUTINE NIOX1
Function: To read the eight coefficients necessary to compute the nitrous oxyde transmission
-1
, step of
10 cm
-1
.
SUBROUTINE NIOX2
Function: Same as NIOX1 but for frequency interval 5060-7610 cm
-1
.
SUBROUTINE NIOX3
-1
.
SUBROUTINE NIOX4
-1
.
SUBROUTINE NIOX5
-1
.
SUBROUTINE NIOX6
-1
.
200
SUBROUTINE OXYG3
Function: To read the eight coefficients necessary to compute the O
2
transmission according
to the Malkmus model (see the text). The frequency interval is 2500-5050 cm
-1
, step of 10 cm
-1
.
SUBROUTINE OXYG4
Function: Same as OXYG3 but for frequency interval 10180-12730 cm
-1
.
SUBROUTINE OXYG5
-1
.
SUBROUTINE OXYG6
-1
.
201
SUBROUTINE OZON 1
Function: To read the eight coefficients necessary to compute the O
3
transmission according to
the Malkmus Model (see the text). The frequency interval is 2500-5050 cm
-1
, steps of 10 cm
-1
.
202
SUBROUTINE WAVA1
Function: To read the eight coefficients necessary to compute the H
2
O transmission
according to the Goody Model (see the text ). The frequency interval is 2500-5050 cm
-1
, steps of 10
cm
-1
.
SUBROUTINE WAVA2
Function: Same as WAVA1 but for frequency interval 5060-7610 cm
-1
.
SUBROUTINE WAVA3
-1
.
SUBROUTINE WAVA4
-1
.
SUBROUTINE WAVA5
-1.
SUBROUTINE WAVA6
-1
.
203
SUBROUTINE DUST
Function: To read the scattering phase function for the Dust-Like component. Computations
have been performed for 83 phase angles (80 Gauss Angles and 0, 90, 180) and 10 wavelengths,
0.400, 0.515, 0.550, 0.633, 0.694, 0.860, 1.536, 2.250, 3.750 m.
SUBROUTINE OCEA
Function: Same as DUST but for the oceanic component.
SUBROUTINE SOOT
Function: Same as DUST but for the soot component.
SUBROUTINE WATE
Function: Same as DUST but for the water-soluble component.
SUBROUTINE BBM
Function: Same as DUST but for biomass burning model.
SUBROUTINE BDM
Function: Same as DUST but for desertic background model.
SUBROUTINE STM
Function: Same as DUST but for stratospheric model.
Reference:
R.A. Mc CLATCHEY, M.J. BOLLE, K.Ya. KONDRATYEV, "A preliminary cloudless standard
atmosphere for radiation computation", Boulder, Colorado, U.S.A. (1982).
E. P. SHETTLE, Optical and radiative properties of a desert aerosol model. Symposium of
Radiation in the atmosphere, 1 Deepak publishing), pp. 74-77, 1984.
M. KING, HARSHVARDHAN, and ARKING, A, A model of the Radiative Properties of the El
Chichon Stratospheric Aerosol Layer, J. Appl. Meteor., 23, (7), pp. 1121-1137, 1984.
204
SUBROUTINE MIDSUM
Function: To read the midlatitude summer model atmosphere, i.e.pressure (mb), temperature
(K), water vapor and ozone concentrations (g/m
3
) as a function of the altitude (34 levels)
Z = 1 km for 0 < Z
km
< 25
Z = 5 km for 25 < Z
km
< 50
Z = 70, 100 km and (p = 0).
200 220 240 260 280 300
-4 -3 -2 -1 0 1 2 3 4
0
20
40
60
80
100
Log(P [mb] )
Temperature [K]
A
l
t
i
t
u
d
e

[
K
m
]
P
T
Mid. Sum. atm.
-11 -10 -9 -8 -7 -6 -5 -4 -3
-10 -8 -6 -4 -2 0 2
0
20
40
60
80
100
Log(H
2
O dens. [g/m
3
])
Log(O3 dens [g/m
3
])
A
l
t
i
t
u
d
e

[
K
m
]
H
2
O
O
3
Mid. Sum. atm.
Reference:
Mc CLATCHEY R.A., FENN R.W., SELBY J.E.A., VOLZ F.E.and GARING J.S., Optical
properties of the Atmosphere, AFCRL-TR- 71-0279, Enviro. Research papers, No 354, L.G.
HANCOM FIEL Bedford, Mass. U.S.A., 1971.
205
SUBROUTINE MIDWIN
Function: To read the midlatitude winter model atmosphere, i.e., pressure (mb), temperature
3
Z = 1 km for 0 < Z
km
< 25
Z = 5 km for 25 < Z
km
< 50
Z = 70, 100 km and (p = 0).
180 200 220 240 260 280 300
-4 -3 -2 -1 0 1 2 3 4
0
20
40
60
80
100
Log(P [mb] )
Temperature [K]
A
l
t
i
t
u
d
e

[
K
m
]
P
T
Mid. Win. atm.
-11 -10 -9 -8 -7 -6 -5 -4 -3
-10 -8 -6 -4 -2 0 2
0
20
40
60
80
100
Log(H
2
O dens. [g/m
3
])
Log(O3 dens [g/m
3
])
A
l
t
i
t
u
d
e

[
K
m
]
H
2
O
O
3
Mid.Win. atm.
properties of the Atmosphere, AFCRL-TR- 71-0279, Enviro. Research papers, No 354, L. G.
206
SUBROUTINE SUBSUM
Function: To read the subarctic summer model atmosphere, i.e., pressure (mb), temperature
3
Z = 1 km for 0 < Z
km
< 25
Z = 5 km for 25 < Z
km
< 50
Z = 70, 100 km and (p = 0).
180 200 220 240 260 280 300
-4 -3 -2 -1 0 1 2 3 4
0
20
40
60
80
100
Log(P [mb] )
Temperature [K]
A
l
t
i
t
u
d
e

[
K
m
]
P
T
Sub. Sum. atm.
-11 -10 -9 -8 -7 -6 -5 -4 -3
-10 -8 -6 -4 -2 0 2
0
20
40
60
80
100
A
l
t
i
t
u
d
e

[
K
m
]
H
2
O
O
3
Sub. Sum. atm.
Log(O
3
dens. [g/m
3
])
Log(H
2
O dens. [g/m
3
])
Reference:
Mc CLATCHEY R.A., FENN R.W., SELBY J.E.A., VOLZ F.E. AFCRL-TR- 71-0279, Enviro.
Research papers, No 354, L.G. HANCOM FIEL Bedford, Mass. U.S.A., 1971.
207
SUBROUTINE SUBWIN
Function: To read the subarctic model atmosphere, i.e., pressure (mb), temperature (K), water
vapor and ozone concentrations (g/m
3
Z = 1 km for 0 < Z
km
< 25
Z = 5 km for 25 < Z
km
< 50
Z = 70, 100 km and (p = 0).
180 200 220 240 260 280 300
-4 -3 -2 -1 0 1 2 3 4
0
20
40
60
80
100
Log(P [mb] )
Temperature [K]
A
l
t
i
t
u
d
e

[
K
m
]
P
T
Sub. Win. atm.
-11 -10 -9 -8 -7 -6 -5 -4 -3
-10 -8 -6 -4 -2 0 2
0
20
40
60
80
100
Log(H
2
O dens. [g/m
3
])
Log(O
3
dens. [g/m
3
])
A
l
t
i
t
u
d
e

[
K
m
]
H
2
O
O
3
Sub. Win. atm.
Reference:
Mc CLATCHEY R.A., FENN R.W., SELBY J.E.A., VOLZ F.E. and GARING J.S., Optical
208
SUBROUTINE TROPIC
Function: To read the tropical model atmosphere, i.e., pressure (mb), temperature (K), water
3
Z = 1 km for 0 < Z
km
< 25
Z = 5 km for 25 < Z
km
< 50
Z = 70, 100 km and (p = 0).
180 200 220 240 260 280 300
-4 -3 -2 -1 0 1 2 3 4
0
20
40
60
80
100
Log(P [mb] )
Temperature [K]
A
l
t
i
t
u
d
e

[
K
m
]
P
T
Trop. atm.
-11 -10 -9 -8 -7 -6 -5 -4 -3
-10 -8 -6 -4 -2 0 2
0
20
40
60
80
100
A
l
t
i
t
u
d
e

[
K
m
]
H
2
O
O
3
Trop. atm.
Log(O
3
dens. [g/m
3
])
Log(H
2
O dens. [g/m
3
])
Reference:
209
SUBROUTINE US 62
Function: To read the U.S. Standard Atmosphere, i.e., pressure (mb), temperature (K), water
3
Z = 1 km for 0 < Z
km
< 25
Z = 5 km for 25 < Z
km
< 50
Z = 70, 100 km and (p = 0).
180 200 220 240 260 280 300
-4 -3 -2 -1 0 1 2 3 4
0
20
40
60
80
100
Log(P [mb] )
Temperature [K]
A
l
t
i
t
u
d
e

[
K
m
]
P
T
US62 atm.
-11 -10 -9 -8 -7 -6 -5 -4 -3
-10 -8 -6 -4 -2 0 2
0
20
40
60
80
100
A
l
t
i
t
u
d
e

[
K
m
]
H
2
O
O
3
U62 atm.
Log(H
2
O dens. [g/m
3
])
Log(O
3
dens. [g/m
3
])
Reference:
Mc CLATCHEY R.A., FENN R.W., SELBY J.E.A., VOLZ F.E. and GARING J.S., Optical
210
211
MISCELLANEOUS
212
213
SUBROUTINE EQUIVWL
Function: To compute the equivalent wavelength needed for the calculation of the
downward radiation field used in the computation of the non lambertian target contribution
(main.f).
Description: The input is the spectral response of the selected sensor as well as the solar
irradiance spectrum. The output is the equivalent wavelength which is computed by averaging the
spectral response of the sensor over the solar irradiance using increment of 2.5nm,
214
SUBROUTINE PRINT_ERROR
Function: provide centralized error handling for the code and output specific error
messages.
215
SUBROUTINE SPECINTERP
Function: To compute the atmospheric properties at the equivalent wavelength (see
EQUIVWL.f) needed for the calculation of the downward radiation field used in the computation of
the non lambertian target contribution (main.f).
Description: The input is the equivalent wavelength for which coupling between BRDF and
atmosphere is to be computed. Using the actual aerosol model, the ouput are the atmospheric
properties at this wavelength that is the aerosol optical thickness, the aerosol single scattering
albedo and the aerosol phase function. In addition, this routine modifies the thickness and single
scattering albedo of the aerosol layer below an aircraft.
216
SUBROUTINE SPLIE2,SPLIN2,SPLINE,SPLINT
Function: Perform interpolation of a furnished BDRF discrete dataset (BRDFGRID.f,
option 1 of BRDF model) to compute the albedo (
=
) as well as the BRDF value at the gaussian
quadrature points.

6s Manual v1.0

Enviado por

Dados do documento

Descrição original:

Direitos autorais

Formatos disponíveis

Compartilhar este documento

Compartilhar ou incorporar documento

Opções de compartilhamento

Você considera este documento útil?

Este conteúdo é inapropriado?

Direitos autorais:

Formatos disponíveis

6s Manual v1.0

Enviado por

Direitos autorais:

Formatos disponíveis

6S User Guide Version 1, November 3, 1995

6S User Guide Version 1, November 3, 1995

6S User Guide Version 1, November 3, 1995

6S User Guide Version 1, November 3, 1995

then we can obtain the percentage density of particles

so for example nj/n for the 3 selected models:

i 2.990 2.990 2.51 2.00

the extinction (or scattering)

b is the scattering coefficient of pigment expressed by

c() and e() are tabular

Você também pode gostar