A comprehensive water quality assessment for a typical river–lake watershed in Northeast China: implications for the water management of boundary lake

作者：Bingbo Ni, Xuemei Liu, Yanfeng Wu, Ming Jiang, Yuanchun Zou · 发表于：Ecological Indicators · 年份：2025 · DOI：10.1016/j.ecolind.2025.113942 · 被引用次数：6 · 研究领域：Water Quality and Pollution Assessment、Hydrology and Watershed Management Studies、Aquatic Ecosystems and Phytoplankton Dynamics

• An improved water quality index (WQI) model was established based on machine learning algorithm. • The WQI condition of 60%−70% monitoring stations in the Muling-Xingkai watershed was good. • River water quality was worse than that of reservoirs and lakes, especially in summer-autumn. • Main driving factors for water quality deterioration differ in summer-autumn and winter-spring. • Nitrogen input and endogenous phosphorus release should be curbed to protect Xingkai Lake. Seasonal freezing and a mismatch between river and lake water quality targets have limited the accurate evaluation of water quality in the northern river–lake system. The water quality of the boundary lake poses a threat to aquatic ecological security and may also affect regional geopolitical stability. Therefore, there is an urgent need for a comprehensive water quality evaluation system to effectively manage the water health of boundary lakes. In this study, we aimed to develop a new comprehensive water quality index model to analyze the water quality status and identify the underlying driving mechanisms within the Muling-Xingkai watershed, thereby proposing effective water management strategies. The XGBoost model and the aggregation function of eight sub-indicators were employed to identify the primary control indicators across various seasons. These methods reduced data redundancy and enhanced the sensitivity of the comprehensive water quality index (WQI) model. The weighted harmonic mean model ( R 2 =...