ESTIMATING FIRE TEMPERATURE WITH SHORT-WAVE INFRARED BANDS OF HIGH RESOLUTION SATELLITES

Soo Chin Liew*

Abstract

This paper describes a novel technique of detecting daytime fires and estimating fire temperature and fire fractional area using the short-wave infrared (SWIR) bands of high resolution satellites such as Sentinel-2, Landsat-9 and WorldView-3. It is a great challenge to detect daytime fires due to the influence of solar reflected radiance. In our algorithm, the surface reflectance is modeled by a linear combination of a small number of reflectance basis functions. The algorithm could then retrieve the sub-pixel fire temperature, fire area, together with the surface reflectance of the background. This technique would enable detection and characterization of small fires in the tropics, where fires are typically small and not as intense as those in the temperate region.

Keywords: Fire, Modeling, Thermal imaging, Basis functions, Sentinel-2

Full Text: