Fire detection in imaging spectrometer data using atmospheric carbon dioxide absorption

Fire detection in shortwave infrared (SWIR) imaging spectrometer data should be possible using the proportion of reflected and emitted radiance absorbed by atmospheric carbon dioxide. A continuum interpolated band ratio (CIBR) was used to measure the relative depth of the 2000 nm carbon dioxide absorption band. Low CIBR values indicated dominantly reflected radiance, while high CIBR values indicated dominantly emitted radiance and the presence of active fire. CIBR values were calculated for Airborne Visible/InfraRed Imaging Spectrometer (AVIRIS) and Hyperion hyperspectral data over fires in California and Arizona. The proposed algorithm has potential uses for fire detection and improving the efficiency of fire temperature retrieval from imaging spectrometer data.