Coupling multi-statistic analysis and 3D-EEM recognition for in-depth pollution evaluation and sources apportionment in a eutrophic shallow lake (Lake Gehu, China)

Abstract

Abstract A clear understanding of both the characteristics and sources of pollution is essential for the rehabilitation and sustainable exploitation of eutrophic shallow lakes. In this study, Lake Gehu was selected as the research subject. Its water quality index (WQI) and trophic level index (TLI) from 2016 to 2020 were evaluated based on multivariate statistical analysis (MSA). Positive matrix factorization (PMF) and three-dimensional fluorescence excitation-emission matrix (3D-EEM) tools were employed to deeply explore the categories and influences of pollution sources. The results indicated improvements to both the comprehensive water quality and eutrophication status with interannual fluctuations over the past five years. However, heavy pollution persisted over summer, with excessive nutrients including organonitrogen, phosphorus and organics. Western and northern Lake Gehu were the most heavily polluted regions, with the significantly higher mean WQI and TLI values compared to those in other regions. Five categories of pollution sources were identified, namely domestic, agricultural non-point, industrial pollution, explosive algal reproduction and the impact of intense rainfall in summer. The priority for control of different pollution sources were presented by PMF, whereas their spatiotemporal influences on Lake Gehu were characterized by 3D-EEM. Thus, coupling PMF and 3D-EEM can enhance the traceability accuracy as well as comprehensively illustrate the characteristics of pollution sources. This study can guide future environmental management of eutrophic shallow lakes.