Spectral mixture analysis of thermal infrared spectrometry data: effects of radiometric resolution on accuracy

At-sensor thermal infrared (TIR) radiation varies with the temperature and emissivity of surface materials, and is affected by atmospheric absorption and emission to varying degrees across the TIR spectrum. One analysis method, temperature emissivity separation (TES) using spectral mixture analysis (SMA), or TESSMA, is extremely dependent on the existence of discrete wavelength bands of sufficient radiometric accuracy to make accurate composition and surface temperature estimates. This analysis focuses on understanding the effects of radiometric inaccuracy on overall estimate accuracy.