정지궤도 기상위성을 이용한 산불탐지 연구

Abstract

Wildfires which can do much damage in ecology, environment, economy and society are recently increasing in frequencies and intensities worldwide. Hence, reliable and continuous monitoring of sudden occurrences of wildfire is demanded, and geostationary meteorological satellites are an alternative to detection of wildfire in large areas. Now, COMS (Communication Ocean and Meteorological Satellite) and MTSAT (Multifunctional Transport Satellite) can observe the area of East Asia but they does not have the official product of wildfire detection. This letter described a wildfire detection algorithm developed for the geostationary satellite that can observe the area of East Asia. Forest fire detection uses brightness temperature in remote sensing. Fire pixels tend to have a significant increase in 4 micrometer brightness temperature but an insignificant increase in 11 micrometer brightness temperature. The algorithm of this study have two major components: (1) to select wildfire candidate pixels and (2) to remove false alarms from the candidates. MODIS(Moderate Resolution Imaging Spectroradiometer), COMS and MTSAT are used for detecting fires in this study. KFS(Korea Forest Service) was used for validating this algorithm in the area of South Korea. The fire product of MODIS was used to compare the result of detection in East Asia. North Korea, China, Japan and Indonesia are study area in East Asia. We adjusted all necessary threshold values using data on April 2011. We then carried out the validation of our algorithm using data on March 2013, the day having the greatest number of forest fire occurrences since the launch of the COMS. We detected various size of wildfires occurred in South Korea on 9 March 2013. This study is the first attempt of wildfire detection using the satellite which can observe the area of East Asia. A lot of case study and quantitative interpretation are needed in East Asia as future works. And retrieval of fire confidence level is required as attribute data of wildfires.