GIS and Remote Sensing

Biography

Biography: Wataru Takeuchi

Abstract

Forest fire has become a global social issue. Originally, forest fills the role of preventing the global warming by photosynthesis. Once it is burned, however, it merely becomes the emission source of carbon dioxide. Besides this one, forest plays a multifunctional role, so wildfire destroying forest has serious effect on global environment and society. For this reason, the damage caused by forest fire must be minimized, and it is necessary to detect forest fire spreading accurately. In this study, the new generation Japanese satellite “Himawari-8” was focused on and forest fire detection was carried out. It is carrying Advanced Himawari-8 Imager (AHI) and the sensor composes of 16 observation bands. We present an approach to evaluate a wildfire duration time with 10 minutes temporal resolution because This extremely high temporal resolution is quite advantageous to understand forest fire spreading. AHI onboard Japanese geostationary satellite imagery is quite powerful to obtain the duration time of rapid fire events such as a grass land fire that cannot be detected with the frequency of Landsat nor MODIS. Research areas are evergreen needleleaf forest in Far-east Russia and evergreen broadleaf forest in Indonesia. Our approach is based on a model that the temperature of the pixel becomes higher than the non-fire pixels if there is some wildfire in the pixel. As a result, it is found that fire duration time is detected by comparing the fire pixel which contains hotspots with a non-fire pixel around it. This technique is useful to detect wildfire duration time even land coverage is evergreen needleleaf forests or evergreen broadleaf forests. We can conclude that an-hourly based monitoring provides us with a sufficient time resolution and plays an important role to monitor wild fire duration time with 10 minutes temporal resolution despite a lower spatial resolution in 2 kilometer than that of MODIS in 1 kilometer.