Solar Influence on Fire Radiative Power Retrieved With the Bispectral Method

• Published in: IEEE transactions on geoscience and remote sensing Vol. 57; no. 7; pp. 4521 - 4528
• Main Authors: Klein, Doris, Richter, Rudolf, Strobl, Christian, Schlapfer, Daniel
• Format: Journal Article
• Language: English
• Published: New York IEEE 01-07-2019; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Algorithms; Bias; Bispectral method; Cameras; Detection; Errors; Evaluation; Fire detection; fire radiative power (FRP); Fires; MODIS; Quantitative analysis; Radiance; Remote sensing; Sensors; Smoldering; Solar power; Solar radiation; solar reflective radiation; Table lookup; TET-1
• ISSN: 0196-2892; 1558-0644
• DOI: 10.1109/TGRS.2019.2891394

Fire radiative power (FRP) is a key product to quantify active fires, which indicates fuel consumption and fire emissions. In the case of the bispectral method, it can be calculated from remote sensing data if a midinfrared (<inline-formula> <tex-math notation="LaTeX">3.8~\mu \text{m} </tex-math></inline-formula>) and thermal infrared channel (<inline-formula> <tex-math notation="LaTeX">\sim 10~\mu \text{m} </tex-math></inline-formula>) are available. While different uncertainty sources have been investigated, the quantitative evaluation of the FRP error as a function of reflected solar radiation is still missing. The ground-reflected solar radiance adds an unknown signal component to the at-sensor radiance during the daytime, which influences the fire detection algorithm as well as the FRP product. FRP errors can reach up to 5%-15% for smoldering fire temperatures of 400-500 K, which is a systematic bias. Errors decrease with increasing temperature and for temperatures higher than 700 K, i.e., flaming fires, the FRP bias is less than 2%. The evaluation is performed for the TET-1 instrument of DLR's FireBIRD mission using the bispectral method.