基于主成分分析的南京市固城湖水质演变特征识别

Identification of the Evolution Characteristics of Water Quality of Gucheng Lake in Nanjing Based on Principal Component Analysis

本文以2016—2021年南京市固城湖水质的逐月监测数据为研究对象,通过主成分分析(PCA)提取了水质的关键信息,并运用Mann-Kendall(M-K)检验及集合经验模态分解(EEMD)对其趋势及周期进行识别。结果表明,pH值、COD_(Mn)、COD_(Cr)、BOD表征的有机质污染物波动对湖区水质变化有着重要影响;近6年来固城湖的综合水质有一定的恶化趋势,特别是pH值、COD_(Mn)、COD_(Cr)、BOD呈现极显著的上升趋势,且呈现显著的季节性差异,其中秋季9月和10月的水质较差;pH值、COD_(Mn)、COD_(Cr)、BOD的主振荡周期为6.5个月,而DO和TP的主周期为9.0个月。PCA、M-K检验、EEMD的组合有效识别了水质的关键时序信息,为湖泊水环境监测及保护提供了重要的数据支撑。

Based on the monthly monitoring data of the water quality of Gucheng Lake in Nanjing from 2016 to 2021,this paper extracts the key information of water quality through principal component analysis(PCA),and the trends and cycles are identified using Mann-Kendall(M-K)test and ensemble empirical mode decomposition(EEMD).The results show that fluctuations in organic matter pollutants represented by pH value,COD_(Mn),COD_(Cr) and BOD have a significant impact on the water quality changes in the lake area.In the past six years,the comprehensive water quality of Gucheng Lake has shown a certain deterioration trend,especially pH value,COD_(Mn),COD_(Cr) and BOD have a significant upward trend,and significant seasonal differences,with poor water quality in September and October of autumn.The main oscillation period of pH value,COD_(Mn),COD_(Cr) and BOD is 6.5 months,while that of DO and TP is 9.0 months.The combination of PCA,M-K test and EEMD effectively identifies the key temporal information of water quality,providing important data support for lake water environment monitoring and protection.