A new algorithm for retrieving land surface temperature and emissivity and applications to airborne hyperspectral SEBASS data

摘要

A new algorithm for retrieving land surface temperature (LST), emissivity and atmospheric parameters (air temperature and total water vapor content) from hyperspectral thermal data is developed. It is based on the multidimensional optimum approach. The surface spectral emissivity is approximated by a Fourier series, and the atmospheric parameters are determined by searching the look-up tables that are created by MODTRAN code. It works reasonably well for the simulated data, and is being applied to airborne hyperspectral SEBASS (Spatially Enhanced Broadband Array Spectrograph System) data.