Modeling directional thermal radiance from a forest canopy

Abstract

Recent advances in remote sensing technology have increased interest in utilizing the thermal-infared region to gain additional information about surface features such as vegetation canopies. Studies have shown that sensor view angle, canopy structure, and percentage of canopy coverage can affect the response of a thermal sensor. These studies have been primarily of agricultural regions and there have been relatively few examples describing the thermal characteristics of forested regions. This paper describes an extension of an existing thermal vegetation canopy radiance model which has been modified to partially account for the geometrically rough structure of a forest canopy. Fourier series expansion of a canopy height profile is used to calculate improved view factors which partially account for the directional variations in canopy thermal radiance transfers. The original and updated radiance model predictions are compared with experimental data obtained over a deciduous (oak-hickory) forest site. The experimental observations are also used to document azimuthal and nadir directional radiance variations. Maximum angular variations in measured canopy temperatures were 4–6°C (azimuth) and 2.5°C (nadir). Maximum angular variations in simulated temperatures using the modified rough surface model was 4°C. The rough surface model appeared to be sensitive to large gaps in the canopy height profile, which influenced the resultant predicted temperature.