10 . Thermal infrared observations of heterogeneous soil - vegetation systems

Heterogeneous soil-foliage systems are characterized under most conditions by large differences in surface temperature between foliage and soil. Two radiometric methods in the thermal infrared spectral region can be used to observe these effects: observations at very high spatial resolution to capture individual leaves or observations at multiple view and illumination angles. The latter is applicable at any spatial resolution, although the anisotropy of emittance is a secondary effect due to thermal heterogeneity and requires high precision radiometric measurements. This chapter reviews the experimental evidence on the anisotropy of emittance by the soil-vegetation system and describes the interpretation of this signal in terms of the thermal heterogeneity and geometry of the canopy space. Observations of the dependence of exitance on view angle by means of ground-based goniometers, airborne and spaceborne imaging radiometers are reviewed first to conclude that under most conditions a two-components, i.e. soil and foliage, model of observed Top Of Canopy (TOC) brightness temperature is adequate to interpret observations. Modeling approaches to describe radiative transfer in the soil-vegetation-atmosphere system, with emphasis on Correspondence/Reprint request: Prof. Dr. Massimo Menenti, D Delft Institute of Earth and Planetary Observation Systems (DEOS), Delft University of Technology, 2629 HS Delft, The Netherlands.. E-mail: [email protected] L. Jia and M. Menenti 2 the thermal infrared region, are reviewed. Given the dependence of observed TOC brightness temperature on leaf-level radiation and heat balance, energy and water transfer in the soil-vegetation-atmosphere system must be included to construct a realistic model of exitance by soil-vegetation systems. Algorithms to estimate foliage and soil temperatures by inverse modeling of multi-angular measurements of exitance are described and evaluated against both synthetic and experimental data.