397 
Figure 8 : Mesured and simulated values of O 0 . 
Inversion method 
Figure 8 shows a good agreement between WSC data and simulated values. We inverted the model to obtain ks 
and sm. In this approach we take into account the temporal variability of surface parameters (soil moisture, 
vegetation and roughness). It can be seen on figure 1 that, soil moisture and vegetation temporal evolutions start 
with low values, increase rapidly from mid July to reach their maximum in August and September respectively; 
then we observed a decrease which corresponds to the end of rainy season. 
1. ) As we do not have enough ct 0 values in April (dry and bare condition), for establishing reference 
values for ks by eliminating soil moisture effect on the signal, we inverted November and December (dry but 
with sparse vegetation (bushes)) data after selecting appropriate soil moisture value. For each incidence angle, 
we obtained ks values which will be used if 0>25° as limiting case in inversion of soil moisture and ks for 
period covering April to August. 
2. ) During this period, soil moisture availability and surface roughness change with rainfall We will alter 
the ks values as derived in 1.) until convergence between simulated and WSC data is reached. The soil 
parameters ks and sm which satisfied this convergence criteria are considered as our reversed parameters.This 
inversion is based on minimization method of rmse (root mean square error) between mesured and simulated 
c 0 . Figure 9 is the temporal evolution of WSC data and soil parameters. 
inversion results 
Figure 9.1 : Temporal evolution of rain, ks and sm(%vol). Figure 9.2 : Temporal evolution of WSC data, ks and sm. 
As we can see on figure 9.1, the roughness parameter ks decreases with time. It is most probably due to soil 
flattening by rain. If we observed figure 9.2, at the beginning of rainy season (up to doy 180) the slope A varies