Assessment of Forest-Stand Destruction by Fires Based on Remote-Sensing Data on the Seasonal Distribution of Burned Area
The paper presents the results of a lethality assessment of forest fires (the proportion of the dead forest areas in comparison to the total area affected by fire) in Russian forests for the period of 2006-2019 according to remote-sensing data. The determined dependence of the probability of pyrogenic forest destruction on the time of the fire onset and its overall impact during the season indicate that the seasonal distribution of the fire area is an informative predictor of lethality. The Seasonal Fire Lethality Index (SFLI) proposed in the article makes it possible to estimate the scale of pyrogenic destruction of Russian forests based on information obtained from satellite monitoring of the area covered by fire distribution during the fire-hazard season. The results demonstrate the presence of a linear correlation between the value of the proposed index and the values of the forest-fire lethality index. At the same time, the determination coefficient of the correlation between the two aforementioned characteristics reaches the maximum (R-2 = 0.80) on July 19; thus, this date was chosen as the climactic day of the forest-fire lethality. The linear regression equation between the characteristics of the fire lethality for Russian forests corresponding to this date allows annual remote assessment with a mean square error of +/- 10.5%. Assessment of the lethality of forest fires based on the proposed method has minimal requirements for input data, which include only information on the distribution of the forest area covered by fire during the fire-hazard season. This makes it possible to use the method on a national scale. The results can be of direct practical importance for the optimization of fire-protection regimes of Russian forests, with an accounting of the potential fire lethality.