Wildfire Monitoring Using Infrared Bands and Spatial Resolution Effects

Abstract

In the past, the application of wildfire monitoring by satellite remote sensing mainly used the mid-infrared band data with kilometer resolution, and there needs to be more quantitative research on the fire detection sensitivity of other infrared bands, like far infrared and short infrared. This study uses the mixed-pixel decomposition method to quantitatively analyze the effect of various spatial resolutions, i.e., 150 m, 300 m, and 1 km resolution infrared bands on fire monitoring. The results show that the fire detection sensitivity of the mid-infrared channel with 150 m resolution is about 30 times higher than that of the 1 km channel. The far-infrared channel with 300 m resolution can detect a fire in hundreds of square meters level; the short-infrared channel with 150 m resolution has an obvious response to the high-intensity flame fire area. This study also uses FY3A meteorological satellite 1 km infrared data and HJ-IB environmental disaster reduction satellite 150 m and 300 m infrared data to verify the above analysis by monitoring individual forest fire cases in Heilongjiang Province in spring and a straw-burning fire in Anhui Province in summer of 2009. The results show that improving the resolution of the infrared band will significantly improve the application ability of satellite remote sensing in small wildfire detection, fire field dynamic monitoring, and fire intensity evaluation.