基于多元统计的西安市河流水质评价及污染源解析

以西安市2009-2017年地表河流水质的逐月监测数据为研究对象,利用多元统计学方法对水质进行综合分析;通过因子分析法探究水中主要污染物及其变化趋势,并分析污染物来源;利用主成分分析法综合评价各个采样点的水质。结果表明:2009-2017年间,24个水质指标可由4~6个主成分描述,方差累计贡献率在78%以上;对西安市河流水质影响最大的指标是有机物和氨氮类;根据主要控制指标的因子载荷,可知2009-2012年间水质整体趋于平稳,2013-2015年间污染加剧,但在2016-2017年又有不同程度的缓解;就采样点的地理位置而言,皂河污染最为严重,而黑河水质整体良好。

Water quality assessment and pollution source analysis of Xi'an river based on multivariate statistics

Taking the monthly monitoring data of surface river water quality in Xi''an from 2009 to 2017 as the research object, the comprehensive analysis of water quality is carried out by using multivariate statistical method, the main pollutants in water and their changing trend are explored by factor analysis method, and the sources of pollutants are analyzed. Comprehensive evaluation of water quality at each sampling point is made by principal component analysis. The results show that 24 water quality indicators can be described by 4 to 6 principal components from 2009 to 2017, and the cumulative contribution rate of variance is over 78%. The most influential indexes of Xi''an river water quality are organic matter and ammonia nitrogen. According to the factor load of the main control indicators, it can be seen that the overall water quality tends to be stable from 2009 to 2012, and the pollution intensifies from 2013 to 2015, but there are different degrees of mitigation in 2016 and 2017. In terms of the geographical location of the sampling points, Zaohe River is the most polluted, while the water quality of Heihe River is good as a whole.