基于机器学习的珠江河口咸潮预测分析

打开文本图片集

中图分类号：TV148 文献标识码：A 文章编号：1001-9235（2025）07-0001-10

Prediction and Analysis of Saltwater Intrusion in Pearl River Estuary Based on Machine Learning

Boheng， ZHANG Jingwen， KANG Zheng*， CHEN Yifan， HUANG Hanliang， LIN Kairong， XIAO Mingzhong （School ofCivil Engineering，Sun Yat-Sen University，Guangzhou51o275，China）

Abstract：Theaccelerationofurbanizationhas inducedasharpincreaseinurban waterconsumption，sothesaltwaterintrusionhas strongerinfluencesondomestic，industrialandagriculturalwateruseinestuarineareas.Toenhancethewatersupplysecurityof coastalcities，itisecssrytoanalyendpredictsaltwaterintrusionTisstudystofurthrinvesgatethifueceofrious factors，includingestuarytidelevel，winddirectionandspeed，andupsreamwaterflow，onthesaltwaterintrusion（chloritynthe estuary）inthePearlRiverestuary，providingscientificsupportforimproving watersuplysecurityincoastalcities.Thestudy determne thelagtimeofdiferentinfluencingfactorsonsaltwaterintrusiontroughcoelationcoeffcntsandconstuctsprdiction modelsforthechloridecontentattheestuariesoftheGuangchangand Pinggang waterpumpingstationsbasedonvarious machine learningmethods toanalyzethesaltwaterintrusioninthe ModaomenwaterwayofthePearlRiverestuary.Thestablishedmodelsshow agodperfomanceinpredictigsalinity.Accordingtotheanalsisofteimportanceofifueningfactors，itisfoundtatustram waterflow hasthegreatest influenceonsaltwater intrusion，followed byestuary tide leveland wind directionandsped.

Keywords： saltwater intrusion;machine learning;Pearl River estuary

磨刀门水道是西江的主要出海口门，其上游西江河段承担多座城市的供水任务，对珠江三角洲的区域水安全有着重要的意义[]。（剩余11429字）