Multivariate statistics as means of tracking atmospheric pollution trends in Western Poland.

This study was carried out over a period of 4 years (2002-2005) at 2 sites located in western Poland differing as regards to human impact by analysis of chemical composition of bulk precipitation. The aim of the study was to determine the sources of pollutions and assess their quantitative contribution to the bulk precipitation composition and to analyse long term-changes in the chemical quality of precipitation. Based on this information the possible transboundary impacts of pollution were also determined. The samples were characterized by determining the values of pH, electrolytic conductivity and concentration levels of Cl(-), F(-), SO(4)(2-), NO(3)(-), Na(+), K(+), Mg(2+), Ca(2+) and NH(4)(+). Analytical measurements were connected with application of principal component regression (PCR) and time series analysis (TS). Based on PCR results three major sources of pollutants in central part of Poland have been identified and quantitatively assessed as follows: "combined" (Poznań - 31%, WNP - 32%), "soil-particulates" (Poznań - 2%, WNP - 26%), "anthropogenic-fossil fuels" (Poznań - 43%, WNP - 23%). Time series analysis enabled discovering 12-month time cycle for NO(3)(-), NH(4)(+), Cl(-), F(-) and SO(4)(2-) in average monthly concentration values in bulk precipitation collected in Wielkopolski National Park. Seasonal variation in the emission of precursors of NO(3)(-) and NH(4)(+) was caused by changes in intensity of fertilizer application in agriculture and automobile exhaust emissions. Decreasing trend was visible for sulphates, nitrates, chlorides and fluorides which is an important indication of the acid rain reduction in the ecologically protected area and in Poznań.