Study of forest fires according to remote sensing data (on the example of the Chornobyl exclusion zone)

Abstract

Summary Remote sensing of the Earth plays an important role in monitoring and assessing the effects of forest fires. With the help of various methods of processing multispectral space images, it is possible to determine the risk of fire spread, identify hot spots and set thermal parameters, map the affected areas and assess the consequences. The aim of the work is to assess the severity associated with the post-fire phase on the example of the forests of the Chornobyl Exclusion Zone. The objectives of the study are to determine the area of burned areas from different time space images obtained from the satellite Sentinel-2 using a normalized burn ratio (NBR). Implemented researches show that the use of remote data of high periodicity and the presence of additional bands of surveying systems significantly expands the range of tasks that can be solved using them. Normalized burned ratio allows quickly and efficiently to identify and calculate the area damaged by fires, that gives possibility operatively assess the consequences of such fires and estimate the damage. It is established that in the studied area the accuracy of area calculation using the normalized burned ratio is 6.7% of the template area, which is sufficient for this type of task.