Statistical Analysis of Emergencies Consequences in Russia

The present study is devoted to the problem of occurrence and development of emergencies in Russia, characterized by high values of the main indicators of the consequences of natural and technogenic emergencies, including fires. Time series models, including trend-seasonal, adaptive and ARIMA-models, were built for the mortality in technogenic fires. The forecast for 2018 was made on the basis of an adaptive model that best approximates the dynamics. The analysis of time series correlation established long-term impact of the dynamic process of fire accidents on the mortality rate in fires. The developed models of population and economic losses by the methods of correlation-regression analysis made it possible to identify the significant risk factors for emergencies inherent in modern Russia. The EM-clustering algorithm was used to classify regions by the level of exposure to catastrophic risks. Typological regressions allowed to detect specific features of certain groups of regions, affecting the death rate in emergency situations. The presented results will help to make effective, mathematically based decisions to reduce the risks of emergencies in Russia.