An approach to forecast the level of air pollution

Abstract

Methodological aspects of analysing the interrelations between air pollution and the main pollution sources by means of statistical modelling are studied. Multifactor regression analysis has been used to describe the air pollution caused mainly by stationary sources. Regression equations for six most widespread pollutants are used to compile air pollution forecasts as a part of long-term planning of environmental protection. The models have been applied on 5 main Estonian towns for the period 1970 to 1995.