Forest Fire Disaster Risk Analysis using Sentinel 2 and Landsat Images Case Study: Al-Qoubaiyat and Tyre regions, Lebanon

Abstract

Fires are considered a threat to the world with all its components and sectors. Recently, it is noticeable an increase in these fires that hit many countries, especially in Lebanon which is considered a country, rich in forests. A forest fire can be naturally caused by either global warming or high temperature. On the other hand, it may be caused by man-made via factories and glass waste. Fires cause great damage to the environment and may lead to human death. Unfortunately, the fire that broke out in AL-Qoubaiyat and Tyre in Lebanon, have been witnessed and caused great damage to the environment, human losses, etc. In this study, a study of fire risk management for those two study areas, will be analyzed using two types of data (Landsat-8 and Sentinel-2) for AL-Qoubaiyat case study, whereas it will be between (Landsat-7 and Sentinel-2) for the Tyre case study. The Analysis will be done by using the Normalized Burn Ration (NBR), Differenced Normalized Burn Ration (NBR) along with all type of required atmospheric corrections. According to our study, it was found advisable to monitor fire risk management using Sentinel-2 L2A data since the atmospheric correction is already performed on it but for L1C data the Sen2Cor python must be used to apply atmospheric correction. Furthermore, the Sentinel-2 L2A data analysis gave more precise results than Landsat-8 by about 2% in Sour case study and 5.7 % in AL-Qoubaiyat case study. Hoping that this method will help in tracking fires, disaster risk reduction, and help in classifying burn severity accompanied with calculating the area corresponding to each class.