首页|基于多源遥感及气象数据的河流非光学活性水质参数反演模型研究

基于多源遥感及气象数据的河流非光学活性水质参数反演模型研究

扫码查看
针对传统河流水质监测成本高、地面监测站点稀疏等问题,基于哨兵2卫星多光谱遥感数据,结合MODIS地表温度、植被指数、气溶胶光学厚度数据产品,以及ERA5气象数据产品中的地表风速数据,以非光学活性参数溶解氧(DO)、化学需氧量(COD)和氨氮(NH3-N)的地面水质监测站的监测数据为参照,采用支持向量回归(SVR)、随机森林(RF)和多层感知机(MLP)3种机器学习方法,通过对比实验选出每种水质参数最优反演模型及其对应的输入特征组合.模型性能测试实验结果表明,利用多源遥感及气象数据反演出的DO、COD以及NH3-N的决定系数(R2)分别为0.896、0.781、0.529,均方根误差(RMSE)分别为 0.263 mg/L、0.383 mg/L、0.061 mg/L.与仅使用哨兵2卫星多光谱遥感数据的反演结果相比,R2分别提高了 7.04%、19.05%、18.34%,RMSE分别降低了 34.58%、37.42%、14.08%.表明多源遥感及气象数据对提高DO、COD以及NH3-N水质参数反演准确性有重要意义.最后,模型鲁棒性评估实验表明,当模型训练数据的代表性与全局数据较为接近时,训练的模型具有较好的时空鲁棒性.
Research on the Retrieval Model of Non-optically Active Water Quality Parameters of Rivers based on Multi-source Remote Sensing and Meteorological Data
In view of the high cost of traditional river water quality monitoring and the sparse ground monitoring stations,based on Sentinel-2 satellite multispectral remote sensing data,combined with MODIS surface tem-perature,vegetation index,aerosol optical thickness data products,and the surface wind speed data in ERA5 meteorological data products,the monitoring data of the surface water quality monitoring stations with non-opti-cal active parameters Dissolved Oxygen(DO),Chemical Oxygen Demand(COD)and ammonia nitrogen(NH3-N)are taken as reference,three machine learning methods,Support Vector Regression(SVR),Ran-dom Forest(RF)and Multilayer Perceptron(MLP),were used to select the optimal inversion model of each water quality parameter and its corresponding input feature combination through comparative experiments.The experimental results of the model performance test show that the determination coefficients(R2)of DO,COD and NH3-N are 0.896,0.781 and 0.529,respectively,and the Root Mean Square Error(RMSE)are 0.263 mg/L,0.383 mg/L and 0.061 mg/L,respectively.Compared with the retrieval results using only Sentinel-2 multi-spectral remote sensing data,R2 increased by 7.04%,19.05%and 18.34%respectively,and RMSE decreased by 34.58%,37.42%and 14.08%respectively.It shows that multi-source remote sensing and meteorological data are of great significance to improve the retrieval accuracy of DO,COD and NH3-N water quality parame-ters.Finally,the model robustness evaluation experiment shows that the trained model has better space-time ro-bustness when the representativeness of the model training data is close to the global data.

Machine learningMulti-source remote sensing dataDissolved oxygenChemical oxygen de-mandAmmonia nitrogenWater quality retrieval model

兑紫璇、王卿、王敏、张璟、顾倩荣

展开 >

中国科学院上海高等研究院低碳转化科学与工程重点实验室碳数据与碳评估中心,上海 201210

中国科学院大学,北京 100049

上海市环境科学研究院,上海 200233

江苏省司法警官高等职业学校,江苏 南京 212008

展开 >

机器学习 多源遥感数据 溶解氧 化学需氧量 氨氮 水质反演模型

上海市科技创新行动计划社会发展科技攻关计划(2020)国家自然科学基金面上项目

20dz120430252178060

2024

遥感技术与应用
中国科学院遥感联合中心

遥感技术与应用

CSTPCD北大核心
影响因子:0.961
ISSN:1004-0323
年,卷(期):2024.39(1)
  • 52