The rationale of this research is to investigate approaches based on modelling and remote sensing data for estimating the spatial distribution of yield and irrigation of wheat in semi-arid areas. The specific objective is to compare the performances of two approaches to test the STICS crop model using remotely sensed estimates of Leaf Area Index (LAI). An experimental study of phenology, yield and water balance of an irrigated wheat was made in the Marrakech-Haouz plain during year 2003. Experimental data allowed to run STICS using two approaches: 1) calibration of the parameters that control the time course of LAI ; 2) driving from LAI time series interpolated with a simple model. The results show the accuracy of STICS to simulate actual evapotranspiration and yield for both approaches. Finally, the two approaches were compared using remotely sensed estimates of LAI upon four scenarios of satellite time revisit frequency. The simulations we obtained always show acceptable results. However, differences appear between the variables, between the approaches and between the frequencies.