基于多源数据和机器学习的洪湖湿地叶绿素a和浊度时空变化研究  

Study on Spatio-temporal variations of Chlorophyll a and Turbidity in Honghu Wetland based on multi-source Data and Machine Learning

作　　者：刘昔 张朦 谢婷婷 曹亮 黄小龙 张磊 王学雷 周正 LIU Xi;ZHANG Meng;XIE Ting-ting;CAO Liang;HUANG Xiao-long;ZHANG Lei;WANG Xue-lei;ZHOU Zheng(Ecological Environment Monitoring and Scientific Research Center,Yangtze River Basin Ecological Environment Supervision and Administration Bureau,Ministry of Ecological Environment,Wuhan 430010,China;Innovation Academy for Precision Measurement Science and Technology,Chinese Academy of Sciences,Wuhan 430061,China)

机构地区：[1]生态环境部长江流域生态环境监督管理局生态环境监测与科学研究中心,湖北武汉430010 [2]中国科学院精密测量科学与技术创新研究院,湖北武汉430061

出　　处：《长江流域资源与环境》2025年第2期374-382,共9页Resources and Environment in the Yangtze Basin

基　　金：湖北省重点研发计划项目(2023BCA003);国家自然科学基金长江水科学研究联合基金项目(U2240213)。

摘　　要：以湖北省第一大湖泊洪湖为研究对象,利用Sentinel-2遥感影像和现场监测数据,基于随机森林机器学习模型构建洪湖湿地叶绿素a和浊度遥感反演模型并揭示其时空变化规律。研究结果表明,叶绿素a与浊度反演模型均能较好地拟合数据,R^(2)的值分别为0.888和0.878;2020~2022年,洪湖叶绿素浓度均值分别为58.127、61.847和82.017μg/L,浊度均值分别为50.180、47.379和85.377 NTU;三年间叶绿素a和浊度显著升高表明洪湖水质处于持续恶化趋势,长期围网围垸养殖过程的迟滞效应、极端气候事件的冲击影响、流域外源污染负荷的居高不下加之内源污染释放的显著增强可能是导致洪湖水质根源性变化的重要驱动因素。该研究揭示洪湖湿地叶绿素a和浊度时空变化规律并分析主要驱动因素,对洪湖水华风险预警和重点区域有效管理具有重要意义。A remote sensing inversion model was developed using a random forest machine learning model to study the spatio-temporal variations of chlorophyll a and turbidity in Honghu Wetland,the largest lake in Hubei Province.The inversion models for chlorophyll a and turbidity showed a satisfactory fit,with R^(2) values of O.888 and 0.878,respectively.Mean chlorophyll concentrations in Honghu Lake increased from 2020 to 2022,with values of 58.127,61.847,and 82.017μg/L.The mean turbidity levels were 50.180 NTU,47.379 NTU,and 85.377 NTU for the same period.The significant increase in chlorophyll a and turbidity over the three years indicated a continuous deterioration in the water quality.Factors such as long-term cage aquaculture,extreme climate events,external pollution load,and endogenous pollution release contributed to these changes.This study provided valuable insights into the spatio-temporal variation of chlorophyll a and turbidity in Honghu Lake to help take early warning and effective management efforts.

关 键 词：多源数据 机器学习 洪湖湿地 时空变化 

分 类 号：X52[环境科学与工程—环境工程] X824