Spatialization of sensible heat flux over a heterogeneous landscape

Two methods for retrieving sensible heat flux over a bare soil/wheat composite surface are compared. The first one is based on field measurements using large aperture scintillometers. The comparison with reference fluxes obtained from eddy correlation technique shows that scintillometry-derived fluxes are overestimated by 10%. A numerical experiment demonstrates this is induced by the non-uniform sensitivity of the scintillometer to C_N² along the path length which follows a 'bell-shaped' curve. The second method is based on the use of a simple surface energy balance model, SEBAL, supplied with high spatial resolution remote sensing data from two airborne sensors in visible, near infrared and TIR bands. A comparison with scintillometry-derived fluxes shows important discrepancies. These result from large errors in the estimation of the roughness length z₀ in the model. This demonstrates that the use of an empirical relationship based on NDVI only is inadequate for inferring this key parameter in SEBAL.