Water quality assessment and contribution rates of main pollution sources in Baiyangdian Lake, northern China

Water quality of Baiyangdian Lake is very important for maintaining ecosystem functions. It is spatially heterogeneous and temporally variable influenced by multiple factors. However, the evolution mechanism of water quality in the lake remains unclear. In this study, water quality index (WQI), multivariate statistical methods and absolute principal component score-multiple linear regression (APCS-MLR) model were applied to assess water quality of the lake and parse main pollution sources. One year (2017–2018) dataset of seven water quality parameters measured in 28 sample sites were analyzed. The results indicated that water quality in summer was relatively poor, and WQI values were relatively high in autumn. Discriminant analysis showed that both standard mode and stepwise mode correctly classified >86.5% of cases. According to results of principal component analysis/factor analysis, three principal components in cluster 1 and cluster 2 in four seasons could explained >80.5% of the total variances, respectively. Besides, water quality variations were mainly related to natural factor, sediment release, domestic sewage, livestock breeding, rainfall runoff and decomposition of reed and lotus. Through APCS-MLR model, the order of contribution of domestic sewage to COD was autumn>spring>summer>winter. Spatially, it was western region>eastern region and northern region>southern region, respectively. Nutrients mainly came from sediment release, rainfall runoff and decomposition of reed and lotus. The average contribution rates of the pollution sources were between 20% and 31%, respectively. This study systematically found the tempal-spacial deterioration degree and main influence factors of water quality, and the contribution rates of main pollution sources to different lake areas in four seasons. The outcomes provide the theory and methodology basis for water environment management. • Spatial heterogeneity of water quality was quantified in each season. • The WQI values of southern region in autumn were relatively high. • TP was identified as one of main discriminant variables in each season. • Nutrients mainly came from sediment, rainfall runoff and decomposition of plants. • Contributions of pollution sources were quantified in distinct areas of four seasons.