The present work is a preliminary study with the aim of characterize the particle size and the chemical composition of aerosols in the Charqueadas County in Rio Grande do Sul State, Brazil, in order to better identify the sources responsible for alterations in the air quality. Seven filters were used for collecting material on thin nucleopore polycarbonate membrane (0.4 μm) during March / 1996 to July / 1996. X-ray energy dispersive electron microscopy was used to determine the chemical composition of aerosol particles, and 21 elements were then analyzed. Particle size distribution was measured by SEM-EDX analysis, and also by automated image technique. Nonhierarchical cluster analysis was applied to identify the types of particles present in the samples. This procedure resulted in the definition of 8 groups of particles containing Fe, Zn, Si, Al, S, Ca, Na and K, that revealed the chemical heterogeneity of aerosols in Charqueadas Country. The particle size analysis showed the predominance (about 80% of all analyzed particles) of the fraction ≤ 10 μm (day) with the highest concentration of these particles located in the size ≤ 2.0 μm (day). Data from particle size distribution, including Falk and Ward parameters, were combined with meteorological parameters and subjected to Hierarchical Cluster Analysis and Discriminant Analysis. This procedure allowed particle-based features and meteorological variables to be integrated for potentially discriminant capability. The particle size analysis showed the predominance (about 80% of all analyzed particles) of the fraction ≤ 10 μm (day) with the highest concentration of these particles located in the size ≤ 2.0 μm (day). Data from particle size distribution, including Falk and Ward parameters, were combined with meteorological parameters and subjected to Hierarchical Cluster Analysis and Discriminant Analysis. This procedure allowed particle-based features and meteorological variables to be integrated for potentially discriminant capability. The particle size analysis showed the predominance (about 80% of all analyzed particles) of the fraction ≤ 10 μm (day) with the highest concentration of these particles located in the size ≤ 2.0 μm (day). Data from particle size distribution, including Falk and Ward parameters, were combined with meteorological parameters and subjected to Hierarchical Cluster Analysis and Discriminant Analysis. This procedure allowed particle-based features and meteorological variables to be integrated for potentially discriminant capability
