SPATIAL MODELING OF MORTALITY AND AIR CONTAMINATION IN THE REPUBLIC OF BASHKORTOSTAN

The aim of the was to study quantitative effects of chemical pollution of atmospheric air by gaseous, liquid, solid substances on the overall population mortality in the Republic of Bashkortostan using mathematical models with spatial lags. Methods. The data were obtained from the official statistics from the Territorial Body of the Federal State Statistics Service for the Republic of Bashkortostan (table C 51 "Distribution of deceased by sex, age groups and causes of death"). The data from 54 municipalities and 21 towns of Bashkortostan from 2002 to 2015 were analysed. Spatial regression methods were applied. Moran and Geary coefficients were used to estimate the presence of spatial autocorrelation links; Spatial autoregressive (SAR) models were used to estimate the effect of emissions on mortality rates. Results. The best fit was provided by the SAR model, according to which, an increase in the number of emissions per 1 ton in a municipality is associated with an increase in the mortality for men by 2,33 cases per 1000. The corresponding data for the total population was 2,02 cases per 1000. However, according to the spatial lag (SDM model), no statistically significant effects on mortality was detected. Conclusion. Results of spatial modeling are in line with the findings on direct associations between gaseous and liquid emissions and total mortality and mortality among men in the same municipality. At the same time, we did not find the evidence indirect spatial influence on mortality.

air pollution, modeling of the mortality rate, spatial autoregression