Coupling canopy functioning and radiative transfer models for remote sensing data assimilation

Crop functioning models (CFM) are used in many agricultural and environmental applications. Remote sensing data assimilation appears as a good tool to provide more information about canopy state variables in time and space. It permits a reduction in the uncertainties in crop functioning model predictions. This study presents the first step of the assimilation of optical remote sensing data into a crop functioning model. It consists in defining a coupling strategy between well known and validated crop functioning and radiative transfer models (RTM), applied to wheat crops. The radiative transfer model is first adapted to consistently describe wheat, considering of four layers in the canopy that contain different vegetation organs (soil, yellow leaves and senescent stems, green leaves and stems, green and senescent ears). The coupling is then performed through several state variables: leaf area index, leaf chlorophyll content, organ dry matter and relative water content. The relationships between the CFM outputs (agronomic variables) and RTM inputs (biophysical variables) are defined using experimental data sets corresponding to wheat crops under different climatic and stress conditions. The coupling scheme is then tested on the data set provided by the Alpilles–ReSeDA campaign. Results show a good fitting between the simulated reflectance data at top of canopy and the measured ones provided by SPOT images corrected from atmospheric and geometric effects, with a root mean square error lower than 0.05 for all the wavebands.