The version 1.0 is disposable to the Sparc/SUN, PC/IBM and 
Risc/IBM systems and can treat images with homogeneous or 
constant atmospheric conditions. It means, to the calculating 
point-of-view, to have a unique set of atmospheric parameters 
to the entire image. 
The SCORADIS 1.0 has three principal modules, which are: 
(a) Data processing in the 5S Model 
This module aims to simplify the data input by keyboard, 
letting the 5S Model “transparent” to the user. The 5S model 
has a semi-rigid input format related to the FORTRAN 
language characteristics. This module can perform atmospheric 
simulations, which is one of the original applications of the 5S 
Model. It is not necessary, using this module, to know all the 
particularities and characteristics of the 5S Model. 
(b) Calculation of look-up tables 
This module is a separated part of the correction procedure. It 
allows the execution of remote correction of images using only 
look-up tables. The correction tables are calculated using the 
atmospheric parameters of the area and transferred to the user 
to correct its images. 
(c) Image processing 
This is the main module of the system that has motivated its 
development. The correction of surrounding effects (1st part) 
and the correction of the “pure” atmospheric effects (2nd part) 
are the two principal parts of the correction procedure. After 
the first treatment, the pixels have the influence of the “pure” 
atmospheric process (scattering and absorption) but no have 
the influence of their surrounding pixels. The subdivision of 
the correction procedure allows to evaluate the influence of 
each module in the final results. It is possible to use only the 
second module when the resolution of the image pixels is 
greater than 100x100m, simplifying the correction process. The 
specific treatment of the surrounding effects is an interesting 
characteristic of the System because they are always neglected 
due to the difficulty to describe them adequately. The grey 
level of the output images can represent radiance or reflectance 
levels. 
The hypothesis of atmospheric homogeneity is very reasonable 
when working with Landsat-TM and Spot-HRV images. For 
Noaa-AVHRR images, it is important to be prepared to work 
with the spatial variability of the atmospheric parameters. The 
processing of images with spatial variability is an important 
improvement of the SCORADIS that will be present in the 
version 2.0 on June 1996. It will use a man-machine interface 
based on X-Windows systems to simplify its use by any kind of 
user. New versions will substitute the version of the 5S Model 
used in the SCORADIS 1.0. 
3. INPUT DATA 
The principal input data that are necessaries to run the 5S 
Model, used by the SCORADIS, are the atmospheric model of 
gaseous components, the type of the aerosols and the 
concentration of the aerosols. Other data like the spectral 
conditions of the satellite bands and the calibration 
coefficients are inside the system. The SCORADIS calculates 
by itself the geometrical conditions of illumination by the sun 
and observation by the satellite. 
We have adapted a spectroradiometer LI1800/LICOR to 
measure the direct solar radiation with a black tube that limits 
its field of view from 180° to 2.4°. The great advantage of this 
adaptation is the possibility to have spectral direct solar 
radiation data from 330nm to 1100nm each 1, 2 or 5nm. 
3.1. Atmospheric model of gaseous components 
The atmospheric model of gaseous components is defined in 
the 5S model most precisely by radiosonde data or 
approximately by the water vapor and ozone contents using a 
standard atmospheric model. 
It was possible to launch some radiosondes only in the 
beginning of the work (June and July 1991) when it was 
observed a good agreement between the tropical atmospheric 
model (McClatchey et allii, 1971) and the experimental data. 
Considering the difficulty to obtain radiosonde data with a 
good spatial and temporal resolution, we define the 
atmospheric model of the gaseous components using the 
tropical model proposed by McClatchey et allii (1971) with the 
water vapor and ozone contents determined experimentally. It 
is justified because, for our purposes, the quantity of water 
vapor and ozone is more important than their distribution 
profile through the atmosphere. This approach is becoming 
more attractive to us due to disposable of precipitable water 
and ozone contents data from TOVS data in our laboratory 
since December 1994. 
3.1.1. Water vapor. The estimate of atmospheric columnar 
water vapor is a matter that interests many investigators it has 
a long time. This is a very important parameter considering the 
high quantity of water vapor presents in the tropical 
atmosphere and the important effects of water vapor in the 
remote sensing images. The more simple way to estimate the 
water vapor contents in absence of any other method more 
accurate is using the tropical profile of water vapor corrected 
by ground measurements of water vapor density. This method 
was used in the first years of our work. 
We are working now in a more accurate approach based on the 
main ideas of the method of differential solar transmission 
measurement presented by various authors like Reagan et allii 
(1987) and Holben & Eck (1990), to our conditions of 
equipment and data. 
We have chosen two extreme wavelengths (870nm and 
1026nm) that have not any gaseous influence (that is, total 
gaseous transmistance equal to one) and another one 
intermediate (948nm) that has a strong influence of water 
vapor, using the gaseous transmission calculated by the 5S 
Model. 
The well-known Equation 1 expresses the relation between the 
gaseous transmitance by the water vapor (£g), the water vapor 
contents (uw) and the air mass (m): 
ug = exp[-k.uw^.m*] (1) 
with k, b and c constants, where k depends on the measurement 
equipment, b and c are near 0.5. Applying the natural 
832 
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B7. Vienna 1996 
logari 
obtair 
The 5 
and m 
to obt 
obtain 
and 1° 
The li 
870nn 
radiati 
atmos; 
at 948 
and th 
at 948; 
. Applyi 
Equati 
of the 
The IN 
CIME] 
This n 
measu 
which 
variabi 
vapor : 
images 
3.1.2. 
propos 
agreen 
The U 
ozone 
allii (1 
This e 
data th 
of atm: 
3.2. At 
We cal 
Law a 
ground 
us. 
Anothe 
methoc