2. BACKGROUND 
The backscattering coefficient (a 0 ) measured by scatterometer depends on viewing conditions (incidence angles 
in our case). It is also function of surface characteristics. As shown by many studies [11] the a ngny 
dependence of a 0 is linked to surface roughness, soil moisture, and biomass. For wet soil, the roughness effect 
is important (surface scattering); while for dry soil and vegetation the signal penetrates the medium (volume 
scattering). It was established that at low incidence angles, ct 0 increases with soil moisture and decreases with 
roughness. However at high incidence angles a 0 increases with biomass and surface roughness. Another 
feature, is the decrease of a 0 with incidence angle. This slope is an inverse function of roughness. To reduce the 
complexity of ct 0 behavior in term of surface parameters, it is useful to select incidence angles for a given 
frequency and polarization. 
3. STUDY AREA 
The study area is a 50*50Km square around BANIZOUMBOU (13 o 31’08“N-02°39’37”E) and correspond to 
the East central site of Hapex Sahel experiment [6]. Representative of the Sahelian zone, this site is 
characterised by a rainy season from June to September. The mean annual rainfall is about 550mm with a strong 
spatial and temporal variability. The vegetation is sparse and its growth is governed by water availability. 
4. DATA 
4.1. Ground data 
They concerned volumetric soil moisture content which were measured during Hapex Sahel experiment [13] by 
INRA team over the East central site. As we work at low frequency (5.4 Ghz), we use measurements on the 0-5 
cm depth. We take advantage of daily rainfall data (1992) acquired by EPSAT [14] on the study area. Fig.l 
shows a temporal evolution of part of these data. 
figure 1 : Temporal evolution of surface parameters. Rainfall (x), soil moisture (*), MSAVI (o). 
4.2. Sattelite data 
42.1. NOAA 11/AVHRR (l.lKm of resolution): We used the visible (0.58-0.68p.) and near infrared 
(0.725-l.lp) daily data of this satellite to express vegetation cover from April to October 1992. The vegetation 
index used is the MSAVI (Modified Soil Adjusted Vegetation Index, [2]) which is computed as: 
MSAVI=(p2-pi)*(l+L)/(p 2 + Pl +L) (1) 
L=l-2.12*(p 2 -p 1 )(p 2 -1.06*p 1 )/(p I +p 2 ) (2)