Evaluation of temperature and emissivity separation method using the hyperspectral data for contrast emissivity surfaces

Abstract

Land surface temperature and emissivity separation (TES) is a key parameter in the physical processes of land surface energy and water balance at regional and global scales. Various methods have been proposed to retrieve the temperature and emissivity for the high emissivity (close to 1) materials. This work addressed the iterative spectrally smooth temperature-emissivity separation method (ISSTES) proposed by Borel (1998) for retrieval of temperature and emissivity from the simulated hyperspectral thermal infrared (TIR) data for contrast (high- and low-) emissivity materials. The results show that small modeling errors less than 0.3 K for temperature and 0.01 for contrast materials are shown in ISSTES algorithm. A sensitivity analysis is carried out and the experimental results show that the instrumental noise, the atmospheric downwelling radiance and the atmospheric transmittance have a great influence on the retrieval accuracy, especially for low-emissivity materials.