Chemometric characterization of the quality of the atmospheric environment in Hong Kong

Abstract

In this paper, chemometric approaches based on correlative analysis, hierarchical clustering, classical and robust principal component analysis were investigated to characterize the quality of the atmospheric environment in Hong Kong. Since the weather behaved quite atypically as concerns both the rains and temperature during the sampling period, it is difficult to clarify the relationship between the chemical species, weather and the environmental quality with the raw data directly. However, with the help of suitable chemometric methods, the correlationship between samples or variables could be investigated. Moreover, pattern recognition based on hierarchical clustering, classical and robust principal component analysis could provide a visual aid for identifying the homogeneity and/or differences of the samples under study. The successful applications to 55 coarse and fine particulate observations from the atmospheric environment in Hong Kong showed that chemometric approaches could develop a clear picture concerning the correlation between samples or variables and especially the linkage of the chemical components, weather and the environmental quality for the atmospheric system with much complexity. © 2005 Elsevier Ltd. All rights reserved.