基于聚类分析和水质标识指数的水质评价方法

An approach for water quality assessment based on hierarchical cluster analysis and comprehensive water quality identification index

为探讨一种适用于大尺度、多断面和长时间的水质评价方法,用层次聚类分析将2006~2008年海河5个监测断面的165个水质样本分为20组,并用方差分析验证了结果的可靠性。用综合水质标识指数表示水质样本的综合特征,与单因子评价法相比,该方法充分考虑了最差水质因子的影响,避免了单因子评价法以偏概全的缺陷,且分辨率高。将评价结果分解到各断面,结果表明:密云水库、官厅水库和于桥水库的水质处于II^III类水之间,可满足作为饮用水水源水的要求,三岔口断面大部分时间为IV^V类水,张家口八号桥断面全年有6个月以上水质属于劣V类。

In order to find a new approach for water quality assessment which could be useful for the monitoring of large-scale,long-term and many sections,hierarchical cluster analysis was applied to group 165 water quality samples into 20 clusters.The samples generated from five monitoring sections of Haihe River during 3 years（2006～2008）.Analysis of variance（ANOVA） was applied to validate the reliability of the results providing by hierarchical cluster analysis.A comprehensive water quality identification index was put forward to show the synthetic quality of water samples.Compared with the simple factor assessment,it not only considered the effect of the worse factor of water quality,but also avoided the limitation of simple factor assessment method which drew conclusion from one part and was highly discriminative.The results were assigned to the five sections.Water quality of the Miyun reservoir,the Guanting reservoir and the Yuqiao reservoir belonged to class II and III and could meet the need of drinking water resource.But water quality of the Sanchakou section mostly belonged to class IV and V.Water quality of the Bahaoqiao section was worse than class V for more than six months.