Modeling radiative transfer in mixed and row intercropping systems

A radiative transfer model for two-dimensional intercropping systems is presented. Based on the turbid medium analogy, it considers the canopy as a set of contiguous cells characterized by leaf area densities and angle distributions for each species. Radiation interception within a cell is inferred from relationships accounting for light partitioning between species. Because of horizontal heterogeneity, mean exchange coefficients between radiation sources and receivers are computed from the course of several elementary beams within the canopy. Radiative exchange coefficients are defined for both the direct and diffuse components of incident radiation, and for scattered radiation. Finally, the radiative balance of the canopy is solved by using the radiosities method. As a partial validation, the model is tested with reflected and transmitted radiation measurements made on a row intercrop. This consists of two rows of tall maize alternating with two rows of small maize sown 40 days later. Agreement between reflected radiation measurements and simulations is satisfactory. For locally transmitted radiation (i.e. between rows of tall maize and between rows of small maize), discrepancies are seen but no estimation bias is detected. As an example, the model is used to investigate the effect of canopy structure on light partitioning for the case of a simulated row intercropping in tropical conditions. Simulations show that light partitioning is most influenced by the vertical stratification of the species. If the whole canopy is horizontally homogeneous, the row effect is small and vertical leaf distribution dominates. Numerical differences between cloudy- and sunny-day simulations are small, except at sun elevations <45° at midday.