Author(s): Qian Yu; Yongcan Chen; Dejun Zhu; Zhaowei Liu
Linked Author(s): Yongcan Chen, Zhaowei Liu, Qian Yu, Dejun Zhu
Keywords: Eutrophication; Source apportionment; Pudu River; Cluster Analysis; Positive Matrix Factorization
Abstract: To help ultimately to work out water eutrophication control measures, the identification of the key parameters having a significant impact on the concentration of Chla and the source apportionment of these parameters have to be carried out. In this paper, Positive Matrix Factorization (PMF) and the R-style hierarchical Cluster Analysis (CA) are applied in the middle and lower reaches of the Pudu River, a severely polluted river in Southwest China. Since the spatially different reaches showing different characteristics have to be discriminately treated, nine water quality monitoring sites are categorized into three spatial groups based on both the Trophic Level Index (TLI) of each site and the results from the Q-style hierarchical CA before the key parameter identification and the source apportionment. These three groups are as follows: Cluster 1 (C1) containing three upstream sites with the heaviest eutrophication status; Cluster 2 (C2) consisting of three downstream sites with less serious eutrophication status; and Cluster 3 (C3) constituted of three tributary sites with relatively the lightest eutrophication status. The data sets on water quality of 11 parameters in three seasons (dry season, level season and rainy season) were got from three field surveys during 2010~2011. According to the results got using the R-style hierarchical CA, 3 key parameters are identified for the sites in C1 while 6 for C2 and 7 for C3. Then PMF is used to identify four latent pollutant sources and estimate the contribution of the identified sources to the concentration of the key parameters. The results reveal that the sources are different among the spatial divisions: sources of C1 containing industrial effluents, municipal effluents, phosphorus factories and non-point agricultural run-off; sources of C2 comprised of pollution from upper reaches, non-point agricultural run-off, internal sources and domestic wastewater; sources of C3 made up of non-point agricultural run-off, internal sources, industrial effluents, municipal or domestic wastewater. Additionally, the main pollutant sources to the key parameters are different in C1, C2 and C3. Decision makers could pay the most attention to the main sources of the key parameters before making policy.
Year: 2013