Duffour C., Lagouarde J. P., Olioso A., Demarty Jérome, Roujean J. L. (2016). Driving factors of the directional variability of thermal infrared signal in temperate regions. Remote Sensing of Environment, 177, p. 248-264. ISSN 0034-4257.
Titre du document
Driving factors of the directional variability of thermal infrared signal in temperate regions
Duffour C., Lagouarde J. P., Olioso A., Demarty Jérome, Roujean J. L.
Source
Remote Sensing of Environment, 2016,
177, p. 248-264 ISSN 0034-4257
Land surface temperature (LST) is a good indicator of the land surface state. The measurement of LST is however prone to directional anisotropy which may severely affect the interpretation of the measurements if it is not corrected. This study aims at determining and describing the impact of various factors on anisotropy of continuous crops at mid-latitudes. The SCOPE (Soil Canopy Observation, Photochemistry and Energy fluxes) model is used as a data generator of directional anisotropy since it enables exploring a very large range of meteorological, biochemical and geometrical conditions. An original indicator, the standard deviation of anisotropy in principal plane, is used in order to investigate the impact of the tested variables and parameters. We found that anisotropy is, at first order, related to seasonal trends, in relation to the amount of incident radiation and the solar zenith angle. Then the geometrical structure of the canopy modifies the anisotropy (LAI, LADF, hot spot parameter) followed by the coupling between the water status of the soil and the stress of canopy. Wind speed which is known for having a significant impact on temperature level has a very limited influence on anisotropy. An analysis of the amplitude of anisotropy in the principal and perpendicular planes (from -50 degrees to 50 degrees zenith) showed that anisotropy can reach up to 11 degrees C and similar to 3.5 degrees C respectively. The impact of satellite orbit on anisotropy is also discussed and it is found that, given the latitudes and the season, the anisotropy can severely affect measurements. This is particularly true when the satellite measurements are acquired in a configuration close to the solar principal plane, which often occur at low latitude. These results are of great help in the context of developing simple methods which could then be integrated into satellite data processing algorithms.
