FIELD: testing.
SUBSTANCE: invention relates to the field of ecology and can be used for zoning the territory of the city by the level of health hazard of the atmosphere for the population. Full-scale measurements of high-priority chemically hazardous substances in the air are taken at points evenly distributed throughout the territory of the city. The data on the content of chemically hazardous substances in the air in different parts of the city are combined into clusters using the Kohonen neural network model based on SOM maps. The hazard coefficients (HQ) are calculated in the selected clusters for each substance by the formula 
, wherein 
is the average concentration of the ith substance, 
is the reference concentration of the ith substance. For each substance of each cluster, the probability hazard coefficient (pHQ) is calculated by the formula 
, wherein 
is the probability that 
will exceed 1 in the cluster k to which the point belongs, calculated using the Bayes formula. The territory is zoned by the level of health hazard for the population by mapping interpolation fields with visualisation of the spatial distribution of the probability hazard coefficient (pHQ) values using the inverse distance weights method.
EFFECT: higher accuracy and adequacy of assessing the territory by the level of hazard of the atmosphere for the health of the population.
1 cl, 2 tbl
