生态环境学报2024，Vol.33Issue(11)：1737-1747,11.DOI:10.16258/j.cnki.1674-5906.2024.11.008
基于机器学习模型的沿海城市河网水系氨氮质量浓度高分辨率遥感估算
High-Resolution Remote Sensing Estimation of Ammonia Nitrogen Concentrations in Coastal Urban River Networks Based on Machine Learning Models
摘要
Abstract
Efficient monitoring of the spatiotemporal distribution of ammonia nitrogen(NH3-N)in river networks is crucial for managing regional water pollution and ensuring the health of the ecological environment.This study developed a machine learning-based remote sensing inversion model,tailored for large water bodies with significant variations in NH3-N concentrations,using 204 in-situ NH3-N data(ranging from 0.026 to 6.210 mg·L-1)and eight high-quality Sentinel-2 MSI images collected in Guangzhou in 2019.The results indicated that the existing models for NH3-N inversion exhibited low accuracy when applied to water bodies in Guangzhou,whereas models with multi-feature inputs demonstrated relatively better predictive capabilities.To address this,principal component analysis(PCA)was employed for feature dimensionality reduction(BC-FDR)after evaluating over 16000 Sentinel-2 band combinations.This approach was integrated with three machine learning models:Extreme Gradient Boosting(XGBoost),Random Forest(RF),and Support Vector Regression(SVR),to construct an optimized band-feature machine learning method for NH3-N inversion.The BC-FDR_XGBoost model performed the best(rc2=0.6872,σRMSEc=0.617 mg·L-1,σMAEc=0.385 mg·L-1,n=102;rv2=0.5436,σRMSEv=0.438 mg·L-1,σMAEv=0.362 mg·L-1,n=44).Additionally,independent validation using 58 in-situ data points of NH3-N(n=33)and trend analysis(n=25)further confirmed the high accuracy of the BC-FDR_XGBoost model(r2=0.5315,σRMSE=0.459 mg·L-1,σMAE=0.287 mg·L-1).The satellite-derived NH3-N estimates showed strong consistency with the in situ data in both spatiotemporal distribution and trend analyses.The average NH3-N concentration in the Guangzhou river network in 2019 met the Class Ⅲ water quality standard,with significantly higher concentrations observed during the dry season than during the wet season(0.795 mg·L-1 vs.0.552 mg·L-1).During the wet season,NH3-N concentrations were generally lower in the northern and southern regions,and relatively higher in the central area.In contrast,during the dry season,only the Nansha district and certain sections of the mainstream exhibited relatively low NH3-N concentrations.This study provides a reference for the development of city-scale,large-area NH3-N remote-sensing inversion models,contributing to the assessment and management of regional water environments.关键词
氨氮模型/河网水系/特征优化/机器学习/Sentinel-2影像/广州Key words
ammonia nitrogen model/river networks/feature optimization/machine learning/Sentinel-2 imagery/Guangzhou分类
资源环境引用本文复制引用
文妮,王重洋,陈星达,陈水森,周霞,于国荣..基于机器学习模型的沿海城市河网水系氨氮质量浓度高分辨率遥感估算[J].生态环境学报,2024,33(11):1737-1747,11.基金项目
国家自然科学基金项目(41801364) （41801364）
珠海市社会发展领域科技计划项目(2320004000154) （2320004000154）
广州市科技计划项目(2023A04J1536) （2023A04J1536）
广东省科学院打造综合产业技术创新中心行动资金项目(2023GDASZH-2023010101) （2023GDASZH-2023010101）
广东省科学院实施创新驱动发展能力建设专项(2018GDASCX-0403) （2018GDASCX-0403）
广东省水利厅水资源节约与保护专项资金项目(2024) （2024）
