Multivariate statistical study of simultaneously monitored cloud water, aerosol and rainwater data from different...

The present paper deals with data interpretation of monitoring of various atmospheric events (cloud water, aerosol and rainwater) at three different elevation levels at Achenkirch profile in an Alpine valley, Tyrol, Austria (Christlumkopf-1758 m, Christlumalm-1280 m and Talboden-930 m a.s.l.) by the use of principal components analysis. From October 1995 to September 1996 sampling sessions for all sites from the profile and for all events were performed for the major ions NH(4)(+), Na(+), K(+), Ca(2+), Mg(2+), Cl(-), NO(3)(-), SO(4)(2-)-44 cases with eight variables for rainwater; 117 cases with eight variables for cloud water samples and 50 cases with seven variables for aerosol (the major ions as in rain- and cloud water but without magnesium) at any of the elevations. The aim of the multivariate statistical treatment was to extract information about latent factors determining the data structure in all of the cases in order to compare and interpret similarities and dissimilarities with respect to the elevation or the type of the atmospheric event. Four latent factors seem to explain over 85% of the total variance for almost all sites and events but the factors have different identification for the different events or sites (e.g. 'anthropogenic', 'crustal', 'neutralization', 'salt'). Thus, a comparison between sites and between events becomes possible. It was found that cloud water and aerosol events are much more similar with respect to data structure (relevant to emission sources or processes of formation) than the same events and rainwater. Further, the upper sites of the profile (Christlumkopf and Christlumalm) also reveal data structure similarity differing from that of the lowest site Talboden.