Abstract
An operational algorithm is proposed to retrieve soil and foliage component temperatures over heterogeneous land surface based on the analysis of bi-angular multi-spectral observations made by ATSR-2. Firstly, on the basis of the radiative transfer theory in a canopy, a model is developed to infer the two component temperatures using six channels of ATSR-2. Four visible, near-infrared and short wave infrared channels are used to estimate the fractional vegetation cover within a pixel. A split-window method is developed to eliminate the atmospheric effects on the two thermal channels. An advanced method using all four visible, near-infrared and short wave channel measurements at two view angles is developed to perform atmospheric corrections in those channels allowing simultaneous retrieval of aerosol opacity and land surface bi-directional reflectance. Secondly, several case studies are undertaken with ATSR-2 data. The results indicate that both foliage and soil temperatures can be retrieved from bi-angular surface temperatures measurements. Finally, limitations and uncertainties in retrieving component temperatures using the present algorithm are discussed.
Acknowledgments
The senior author is funded by the Netherlands Users Support Program (SRON) under grant no. EO-049. This work was jointly funded by the Netherlands Remote Sensing Board (BCRS), the Netherlands Ministry of Agriculture, Nature Management and Fisheries (LNV), and the Royal Netherlands Academy of Science (KNAW). Zhengming Wan was supported by NASA EOS Program contract NAS5-31370. The meteorological air temperature data were kindly provided by Dr B. J. J. M. van den Hurk of the Royal Netherlands Meteorological Institute (KNMI).