17 
GOES and METEOSAT 
zil (Holben et al., 1994) 
here are plans to expand 
ernational research and 
\alysis of remote sensing 
i of NASA and ADEOS 
EFFECT 
at describe the surface 
rmospherically resistant 
1 for remote sensing of 
duled to be launched in 
at TM, and the SeaWiFS 
■linear gain the SeaWiFS 
takes advantage of the 
, in addition to the red 
t compose the present 
f ARVI to atmospheric 
rrection process for the 
diance between the blue 
47 
) 
ally resistant vegetation 
ation in the red channel 
ation content of p is 
t in the blue channel is 
much smaller than that 
at the reflectance in the 
nospheric effect in this 
tic and natural surface 
similar dynamic range to 
1 times less sensitive to 
ant is much better for 
e to small size aerosol 
rticle size (e.g. maritime 
le red channel is a bias 
T was chosen so that on 
positive and sometime 
ospheric effect but also 
resistance of ARVI, it is 
expected that remote sensing from MODIS of the vegetation index over most land surfaces 
will include molecular and ozone correction with no further need for aerosol correction, 
except for dust conditions, like in the Sahel. 
Fig. 7: The original vegetation 
index, NDVI (dashed lines) and 
the new vegetation index, ARVI 
(solids lines), as a function of the 
fraction of the surface covered by 
vegetation for forest, grass and 
alfalfa. Computations were 
performed using the 5S radiative 
code (Tanre et al., 1990). Solar 
zenith angle is 60° and nadir view. 
(After Kaufman and Tanre, 1992). 
5. Polarization Measurements 
FRACTION OF VEGETATION 
Polarization measurements add an additional dimension to remote sensing. The 
potential of polarization has been demonstrated in characterizing the atmospheres of 
planets with large optical thickness like those of Venus and Saturn (Hansen and Hovenier, 
1974; Santer and Herman, 1979; Dollfus, 1979) or for studying the surface of Mars where the 
atmosphere is very tiny (Egan, 1969). Extensive measurements of the earth-atmosphere 
system were provided by photographs in polarized light taken during the American space 
Shuttle missions (Egan et al., 1989; Roger et al., 1994). Specific experiment was also 
performed from balloon over ocean (Deuze et al., 1989) 
What can we expect from polarization measurements over land? A major concern for 
the use of polarized light over land surfaces is the capability to discriminate between 
polarization generated in the atmosphere and that generated by the surface. In spring 1990, 
an experiment was conducted in the southern part of France over land surfaces (Deuze et 
al., 1993) with the POLDER airborne simulator (Deschamps et al., 1994). The 'La Crau' site, 
in the south part of France over which the instrument flew, is composed of several zones 
with different kinds of vegetation, grass, rice, wheat, sparse grass and bare soils. The land 
cover was therefore non-uniform and displayed radiance contrasts. The surface 
contribution to the polarized radiance was attenuated by the atmosphere. For wavelengths < 
0.7 gm, the measured polarized reflectance at the top of the atmosphere exhibited a smooth 
pattern suggesting that the polarized light is coming mainly from the atmosphere. At 
A.=850nm, the images included high frequency variability indicating the larger relative 
surface contribution. Since this experiment shows that polarized radiance in the green and 
in the blue parts of the spectrum is mainly controlled by the atmospheric processes, the 
satellite signal can be used for retrieving aerosols properties. In near future, polarization 
measurements from satellites will be performed by the POLDER instrument (Deschamps et 
al., 1994) aboard the ADEOS mission of NASD A in 1996 and by the EOSP aboard the EOS-B 
of NASA in 2003.