• 关键词：绝对主成分  多元线性回归  污染源解析  受体模型  高原湖泊

• 基金项目：国家自然科学基金（No.41701631）；云南省科技计划项目重点研发计划（No.2018BC002）

• 杜展鹏
• 云南大学生态学与环境学院, 昆明 650091

• 王明净
• 云南大学生态学与环境学院, 昆明 650091

• 严长安
• 昆明市生态环境科学研究院, 昆明 650032

• 高伟
• 云南大学生态学与环境学院, 昆明 650091

• 摘要：定量解析污染源是湖泊流域水环境管理的重要基础.基于滇池草海和外海多年水质监测数据，采用主成分分析（PCA）方法识别了主要水质指标的污染源类型，利用绝对主成分-多元线性回归模型（APCS-MLR）得到不同污染源对水质的贡献程度.结果表明，草海主要的污染源有农业面源、城市面源和内源3类，外海的主要污染源是农业面源、城镇生活污染源、城市面源和内源4类.与河流水污染源解析结果不同，底泥内源与气象因子对滇池主要水质指标的影响较大.

• Abstract：Identifying pollution sources is an important basis for water environment management in lake basins. Based on multiple years of water quality monitoring data of Caohai and Waihai in the Dianchi Lake, principal component analysis (PCA) method was applied to determine main pollution sources of key water quality index. Contribution of different pollution sources to key water quality index in the Caohai and Waihai Lakes were quantified by using an established absolute principal component score-multiple linear regression model (APCS-MLR). Results show that the main pollution sources of Caohai Lake are urban sewage-agricultural source, urban runoff and inner sources, while those of Waihai Lake are agricultural source, urban sewage, urban runoff and inner sources. Different from river water pollution sources identification, climatic and inner sources have a high contribution to the main water quality indexes of Dianchi Lake, and even become the largest influence factors.

• Key words：absolute principal component  multiple linear regression  source apportionment  receptor model  plateau lake

• 
  摘要点击次数： 1075 全文下载次数： 1402