Spatial variability effects on the modelling of thermal fluxes in a 3D canopy structure including non-radiative energy exchanges

The use of radiative remotely sensed temperature derived from radiance data acquired by airborne and spaceborne sensors remains controversial. While some researchers are quite skeptic about the actual usefulness of such temperature data, other reported results appear to be extremely accurate, with reported errors for temperature retrievals which are difficult to obtain even in laboratory conditions. Because temperature is an intrinsically dynamic variable, and because all remote sensors are is most cases observing 3D structures which are horizontally and vertically heterogeneous in temperature distributions, the concept of "temperature of the system" is even unclear. Spatial variability effects must be realistically described, taking into account non-linearity in radiance measurements, as well as nonradiative effects intervening in energy exchanges. A consistent description of all intervening effects by means of an appropriate description of the 3D structure is the key to understand the radiative temperature measurements over heterogeneous systems. Two main applications are discussed in more detail: the use of radiative temperature data in energy balance calculations, and the use of remotely sensed temperature data in fire detection over vegetated areas. Although no solutions are given to the difficult problem of interpreting temperature measurements over 3D inhomogenous structures, some guidelines are given to help understanding observations and making proper use of the data in realistic cases.