| 摘要: |
| 河流在城市生态系统中扮演着重要角色，但随着城市化加剧，大量未经处理的生活污水和工业废水排放严重恶化了水质。为准确识别和量化长江南通段水质污染因子，本研究采用了绝对主成分分析(APCS)优化输入的多模型回归分析方法，包括多元线性回归(MLR)、岭回归(RR)和套索回归(LR)。结果表明，APCS-RR模型在处理高度相关的水质数据方面表现优异，有利于精准识别和量化不同季节的污染因子贡献；长江南通段水质在枯水期与丰水期呈现显著差异，枯水期以无机盐类和矿物质污染为主，丰水期则以有机物和营养盐污染为主导；加强工业排放管控、优化农业种植模式、实施分季节的水质管理政策和措施等，可有效应对水质季节性差异带来的挑战。研究成果为当地水质与水环境管理策略奠定了数据基础，也为其他地区精确识别水质污染因子提供了可借鉴的方法。 |
| 关键词: 水质污染 污染因子识别 多模型回归 绝对主成分得分 长江南通段 |
| DOI： |
| 分类号: |
| 基金项目:国家重点研发计划(2023YFC3208903)；南通市市级科技计划项目(MS12022010)；国家自然科学(42101384)；江苏省自然科学(BK20210043)；湖南省自然科学(2023JJ30238)。 |
|
| Identification and Analysis of Water Pollution Factors Based on APCS Model Input Optimization |
|
CHEN Xiangyang1, HUANG Guoqing2, LEI Shaohua2, XIAO Weifeng3
|
|
1.College of Architecture and Engineering,Nantong Vocational University;2.National Key Laboratory of Water Disaster Prevention,Nanjing Hydraulic Research Institute;3.School of Earth Sciences and Spatial Information Engineering,Hunan University of Science and Technology
|
| Abstract: |
| Rivers play an important role in urban ecosystems, but with the intensification of urbanization, the discharge of untreated domestic and industrial wastewater has severely deteriorated the water quality. In order to accurately identify and quantify the water pollution factors in the Nantong section of the Yangtze River, this study adopted the optimized input multiple-model regression analysis method, which includes multivariate linear regression (MLR), ridge regression (RR), and lasso regression (LR), using absolute principal component scores (APCS). The results showed that the APCS-RR model performed well in handling highly correlated water quality data, facilitating accurate identification and quantification of pollution factor contributions in different seasons. The water quality in the Nantong section of the Yangtze River exhibited significant differences between dry and wet seasons, with inorganic salts and mineral pollutants dominating during the dry season, and organic matter and nutrient salts dominating during the wet season. Strengthening industrial emission control, optimizing agricultural cultivation patterns, implementing seasonal water quality management policies and measures, etc., can effectively address the challenges posed by seasonal differences in water quality. The research findings provide a data foundation for local water quality and water resource management strategies, and also offer a reference method for accurately identifying water pollution factors in other regions. |
| Key words: water quality pollution pollution factor identification multi-model regression absolute principal component scores the Nantong section of the Yangtze River |
