Assimilation of polarimetric C-band radar data into CERES-Wheat model

This study investigates the potential of synergistically using multi-temporal ENVISAT ASAR data and crop growth models, such as CERES-Wheat, in order to improve the accuracy of wheat yield predictions. The reliability of crop growth models strongly depends on the accuracy of their numerous inputs, which are seldom available at the appropriate spatial resolution. In this respect, remote sensing may provide valuable information able to lead to large improvements in accuracy of spatially distributed crop growth model estimates. Recent experimental studies have suggested the use of HH/VV radar backscatter ratio, acquired at high incidence angles, to retrieve wheat biophysical parameters, such as above ground biomass and leaf area index (LAI). This work presents a methodology to assimilate LAI, retrieved by means of multi-temporal ENVISAT ASAR data at field scale, into CERES-Wheat crop model. The results, obtained over the Matera site (Italy), show that the assimilation leads to significant improvements in wheat dry biomass and grain yield model estimates. Applying this method the average error on the model yield predictions decreases from 18% to 4.2%.