Abstract

Issues related to thermal infrared radiative transfer modelisation and retrieval of component temperatures from directional remote sensing measurements are addressed in this paper. We compare five models and assess their accuracy in direct mode (directional radiative temperature simulations) and inverse mode (soil and leaf temperature retrievals), for a large set of simulations. We write all models using a single pattern, and we propose a unique inversion scheme based on this pattern. In this context, some precisions are given regarding previous parameterizations and results. Two new efficient and simple models appear to successfully compare to the reference model. The relation between radiative temperature and soil moisture is also examined using a coupled process-radiative transfer model and it is shown that directional measurements dramatically improve the soil moisture detection capability in the thermal infrared.