ANNOUNCEMENTS
Forest fire is one of the major disasters affecting forested and other vegetated areas like grasslands worldwide. It can be defined as an uncontrolled, large and fast spreading fire that occurs in a natural vegetation ecosystem such as forests, grasslands and shrublands. They lead to loss of natural vegetation, release of harmful gases in the atmosphere, causes soil erosion and affects the microclimate and the wildlife dependent on the forests. World over several hectares of forest are damaged annually due to fires. In India too, forest fires have become a regular phenomenon destroying the forest wealth. Uttarakhand, a state situated in northern India and part of the central Himalayan landscape is extremely prone to forest fires. In this study forest fire burnt area of Uttarakhand is mapped for the past nine years from 2014 to 2022 using a machine learning algorithm in Google Earth Engine. Active fires dataset of VIIRS has been used to determine the period of occurrence of forest fires. Different driving parameters of forest fire which vary according to time such as climatic parameters like temperature, precipitation and vegetation parameters like Normalized Difference Vegetation Index and Normalized Difference Water Index were taken into account and the statistical relationship between the parameters and the yearly burnt area were quantified. Statistical modelling was done for prediction of burnt area for 2023. The study shows that the area burnt due to forest fire has a cyclical pattern. It was found that burnt areas have a positive relation with temperature and negative relation with the other parameters. Also, it is observed that the parameters are interlinked with one another for forest fire to occur. The burnt area distribution detection, identification of the link between driving factors and forest fire is beneficial for planning mitigation measures and effective management of forest fires.
Keywords: Forest Fire, Burnt Area, Google Earth Engine, VIIRS Active Fires Dataset, Driving Factors.
