Multivariate analysis of .uorescence and source identi.cation of dissolved organic matter in Jiaozhou Bay, China

Hierarchical clustering analysis and principal component analysis (PCA) methods were used to assess the similarities and dissimilarities of the entire Excitation-emission matrix spectroscopy (EEMs) data sets of samples collected from Jiaozhou Bay, China. The results demonstrate that multivariate analysis facilitates the complex data treatment and spectral sorting processes, and also enhances the probability to reveal otherwise hidden information concerning the chemical characteristics of the dissolved organic matter (DOM). The distribution of di.erent water samples as revealed by multivariate results has been used to track the movement of DOM material in the study area, and the interpretation is supported by the results obtained from the numerical simulation model of substance tracing technique, which show that the substance discharged by Haibo River can be distributed in Jiaozhou Bay.