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

doi:10.16258/j.cnki.1674-5906.2024.11.008

生态环境学报 ›› 2024, Vol. 33 ›› Issue (11): 1737-1747.DOI: 10.16258/j.cnki.1674-5906.2024.11.008

• 研究论文【环境科学】 • 上一篇下一篇

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

文妮¹^,²^,³(), 王重洋²^,³^,^*(), 陈星达³, 陈水森³, 周霞³, 于国荣¹^,^*()

1.昆明理工大学电力工程学院，云南昆明 650500
2.广东省科学院珠海产业技术研究院有限公司，广东珠海 519090
3.广东省科学院广州地理研究所，广东广州 510070

收稿日期:2024-05-22 出版日期:2024-11-18 发布日期:2024-12-06
通讯作者: 于国荣。E-mail: YUGuo_rong6314@163.com
*王重洋。E-mail: wangchongyang@gdas.ac.cn;
作者简介:文妮（1999年生），女（彝族），硕士研究生，研究方向为水环境遥感。E-mail: wenn6201@gmail.com
基金资助:
国家自然科学基金项目(41801364);珠海市社会发展领域科技计划项目(2320004000154);广州市科技计划项目(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);广东省科学院实施创新驱动发展能力建设专项(2018GDASCX-0403);广东省水利厅水资源节约与保护专项资金项目(2024)

High-Resolution Remote Sensing Estimation of Ammonia Nitrogen Concentrations in Coastal Urban River Networks Based on Machine Learning Models

WEN Ni¹^,²^,³(), WANG Chongyang²^,³^,^*(), CHEN Xingda³, CHEN Shuisen³, ZHOU Xia³, YU Guorong¹^,^*()

1. School of Electrical Engineering, Kunming University of Science and Technology, Kunming 650500, P. R. China
2. Zhuhai Industrial Technology Research Industrial, Guangdong Academy of Sciences, Zhuhai 519090, P. R. China
3. Guangzhou Institute of Geography, Guangdong Academy of Sciences, Guangzhou 510070, P. R. China

Received:2024-05-22 Online:2024-11-18 Published:2024-12-06

摘要/Abstract

摘要：

高效监测河网水系氨氮（NH₃-N）的时空分布对区域水体污染防控治理和生态环境健康发展具有重要意义。基于2019年在广州市收集的204个NH₃-N实测数据（0.026－6.210 mg·L^-1）和8景高质量Sentinel-2 MSI遥感影像，发展了适用于大范围水域、NH₃-N质量浓度差异显著的机器学习遥感反演模型。结果显示，已有的NH₃-N反演模型应用于广州市水体时精度受限，但多特征输入的模型预测能力相对较好。在检索16000多种Sentinel-2波段组合的基础上，利用主成分分析方法进行了特征降维（BC-FDR），并结合极端梯度提升（XGBoost）、随机森林（RF）、支持向量回归（SVR）3种机器学习模型，构建了波段特征优化的机器学习NH₃-N反演方法。其中BC-FDR_XGBoost模型表现最佳（r_c²=0.6872，σ_RMSEc=0.617 mg·L^-1，σ_MAEc=0.385 mg·L^-1，n=102；r_v²=0.5436，σ_RMSEv=0.438 mg·L^-1，σ_MAEv=0.362 mg·L^-1，n=44）。另外基于58个实测数据进行了独立验证（n=33）和趋势检验（n=25），结果进一步表明，BC-FDR_XGBoost模型的精度较高（r²=0.5315，σ_RMSE=0.459 mg·L^-1，σ_MAE=0.287 mg·L^-1），卫星遥感反演结果与实测数据在时空分布和变化趋势上具有良好的一致性。2019年，广州市河网水系NH₃-N质量浓度平均为Ⅲ类水质等级，枯水期（0.795 mg·L^-1）显著高于丰水期（0.552 mg·L^-1）。空间上，丰水期NH₃-N质量浓度整体呈南北部低、中部相对较高的特点；枯水期仅南沙区及部分干流NH₃-N相对较低。该研究为建立城市尺度大区域范围水体NH₃-N遥感反演模型提供了参考，有助于区域水环境的评价和治理。

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

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 (r_c²=0.6872, σ_RMSEc=0.617 mg·L^-1, σ_MAEc=0.385 mg·L^-1, n=102; r_v²=0.5436, σ_RMSEv=0.438 mg·L^-1, σ_MAEv=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, σ_RMSE=0.459 mg·L^-1, σ_MAE=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.

Key words: ammonia nitrogen model, river networks, feature optimization, machine learning, Sentinel-2 imagery, Guangzhou

中图分类号:

X522

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.

WEN Ni, WANG Chongyang, CHEN Xingda, CHEN Shuisen, ZHOU Xia, YU Guorong. High-Resolution Remote Sensing Estimation of Ammonia Nitrogen Concentrations in Coastal Urban River Networks Based on Machine Learning Models[J]. Ecology and Environment, 2024, 33(11): 1737-1747.

图/表 17

参考文献 36

[1]	CAO X, ZHANG J, MENG H B, et al., 2023. Remote sensing inversion of water quality parameters in the Yellow River Delta[J]. Ecological Indicators, 155: 110914.
[2]	CHAWLA I, KARTHIKEYAN L, MISHRA A K, 2020. A review of remote sensing applications for water security: Quantity, quality, and extremes[J]. Journal of Hydrology, 585: 124826.
[3]	CHEN J Y, CHEN S S, FU R, et al., 2021a. Remote sensing estimation of chlorophyll-A in case-II waters of coastal areas: Three-band model versus genetic algorithm-artificial neural networks model[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 14: 3640-3658.
[4]	CHEN B T, MU X, CHEN P, et al., 2021b. Machine learning-based inversion of water quality parameters in typical reach of the urban river by UAV multispectral data[J]. Ecological Indicators, 133: 108434.
[5]	CHEN P, WANG B, WU Y L, et al., 2023. Urban river water quality monitoring based on self-optimizing machine learning method using multi-source remote sensing data[J]. Ecological Indicators, 146: 109750.
[6]	DONG G Q, HU Z Q, LIU X, et al., 2020. Spatio-Temporal variation of total nitrogen and ammonia nitrogen in the water source of the middle route of the south-to-north water diversion project[J]. Water, 12(9): 2615.
[7]	DÖRNHÖFER K, OPPELT N, 2016. Remote sensing for lake research and monitoring-Recent advances[J]. Ecological Indicators, 64: 105-122.
[8]	EJAZ U, KHAN S M, JEHANGIR S, et al., 2024. Integrated analytics and machine learning for water quality index assessment-Monitoring of industrial waste polluted stream[J]. Journal of Cleaner Production, 450: 141877.
[9]	HAN D, CURRELL M J, CAO G, 2016. Deep challenges for China’s war on water pollution[J]. Environmental Pollution, 218: 1222-1233.
[10]	KUAN H F, LI J, ZHANG X J, et al., 2020. Remote estimation of water quality parameters of medium-and small-sized inland rivers using Sentinel-2 imagery[J]. Water, 12(11): 3124.
[11]	LI N, NING Z Y, CHEN M, et al., 2022. Satellite and machine learning monitoring of optically inactive water quality variability in a tropical river[J]. Remote Sensing, 14(21): 5466.
[12]	LIU B, XI H J, LI T H, et al., 2023. Black-odorous water bodies annual dynamics in the context of climate change adaptation in Guangzhou City, China[J]. Journal of Cleaner Production, 414(11): 137781.
[13]	LIU Y, WU X, QI W C, 2024. Assessing the water quality in urban river considering the influence of rainstorm flood: A case study of Handan city, China[J]. Ecological Indicators, 160: 111941.
[14]	LO Y, FU L, LU T C, et al., 2023. Medium-sized lake water quality parameters retrieval using multispectral uav image and machine learning algorithms: A case study of the yuandang lake, China[J]. Drones, 7(4): 244.
[15]	PAHLEVAN N, CHITTIMALLI S K, BALASUBRAMANIAN S V, et al., 2019. Sentinel-2/Landsat-8 product consistency and implications for monitoring aquatic systems[J]. Remote sensing of Environment, 220: 19-29.
[16]	SAGAN V, PETERSON K T, MAIMAITIJIANG M, et al., 2020. Monitoring inland water quality using remote sensing: Potential and limitations of spectral indices, bio-optical simulations, machine learning, and cloud computing[J]. Earth-Science Reviews, 205(3): 103187.
[17]	SEO J Y, WON J G, LEE H S, et al., 2024. Probabilistic monitoring of meteorological drought impacts on water quality of major rivers in South Korea using copula models[J]. Water Research, 251: 121175.
[18]	SHI X S, QIU Z F, HU Y J, et al., 2024. Monitoring the water quality distribution characteristics in the Huaihe River Basin based on the Sentinel-2 Satellite[J]. Water, 16(6): 860.
[19]	SONG W, YINGLAN A, WANG Y T, et al., 2024. Study on remote sensing inversion and temporal-spatial variation of Hulun lake water quality based on machine learning[J]. Journal of Contaminant Hydrology, 260: 104282.
[20]	TANG W Z, PEI Y S, ZHENG H, et al., 2022. Twenty years of China’s water pollution control: Experiences and challenges[J]. Chemosphere, 295: 133875.
[21]	TIAN D, ZHAO X F, GAO L, et al., 2024. Estimation of water quality variables based on machine learning model and cluster analysis-based empirical model using multi-source remote sensing data in inland reservoirs, South China[J]. Environmental Pollution, 342: 123104.
[22]	TIAN S, GUO H W, XU W, et al., 2023. Remote sensing retrieval of inland water quality parameters using Sentinel-2 and multiple machine learning algorithms[J]. Environmental Science and Pollution Research, 30(7): 18617-18630.
[23]	TOMING K, KUTSER T, LAAS A, et al., 2016. First experiences in mapping lake water quality parameters with Sentinel-2 MSI imagery[J]. Remote Sensing, 8(8): 640.
[24]	TOPP S N, PAVELSKY T M, JENSEN D, et al., 2020. Research trends in the use of remote sensing for inland water quality science: Moving towards multidisciplinary applications[J]. Water, 12(1): 169.
[25]	WANG C Y, CHEN S S, LI D, et al., 2017. A Landsat-based model for retrieving total suspended solids concentration of estuaries and coasts in China[J]. Geoscientific Model Development, 10(12): 4347-4365.
[26]	WANG C Y, LI W J, CHEN S S, et al., 2018. The spatial and temporal variation of total suspended solid concentration in Pearl River Estuary during 1987-2015 based on remote sensing[J]. Science of the Total Environment, 618: 1125-1138.
[27]	WANG C Y, WANG D N, YANG J, et al., 2020. Suspended Sediment within Estuaries and along Coasts: A Review of Spatial and Temporal Variations based on Remote Sensing[J]. Journal of Coastal Research, 36(6): 1323-1331.
[28]	WANG J B, ZHOU W Q, ZHAO M L, et al., 2023. Water quality assessment and pollution evaluation of surface water sources: The case of Weishan and Luoma Lakes, Xuzhou, Jiangsu Province, China[J]. Environmental Technology & Innovation, 32: 103397.
[29]	WANG X L, FU L, HE C S, 2011. Applying support vector regression to water quality modelling by remote sensing data[J]. International Journal of Remote Sensing, 32(23): 8615-8627.
[30]	ZHANG H W, XUE B L, WANG G Q, et al., 2022. Deep learning-based water quality retrieval in an impounded lake using landsat 8 imagery: An application in Dongping lake[J]. Remote Sensing, 14(18): 4505.
[31]	广州市生态环境局. 2019年广州市环境质量状况公报[EB/OL]. [2020-06-03]. https://sthjj.gz.gov.cn/zwgk/hjgb/index_5.html.
	Guangzhou Municipal Ecological Environment Bureau. 2019 Guangzhou Environmental Quality Report[EB/OL]. [2020-06-03]. https://sthjj.gz.gov.cn/zwgk/hjgb/index_5.html.
[32]	高佳欣, 林昱坤, 涂耀仁, 等, 2023. 遥感反演技术应用于监测地表水体水质参数的现状与展望[J]. 遥感信息, 38(6): 1-14.
	GAO J L, LIN Y K, TU Y R, et al., 2023. Current status and prospects of remote sensing inversion technology applied to monitoring water quality parameters of surface water[J]. Remote Sensing Information, 38(6): 1-14.
[33]	郭荣幸, 王超梁, 陈济民, 等, 2024. 基于Sentinel-2多光谱遥感影像的小浪底水质反演[J]. 人民黄河, 46(1): 93-96.
	GUO R X, WANG C L, CHEN J M, et al., 2024. Remote sensing inversion of water quality of Xiaolangdi Reservoir based on Sentinel-2 Multi-Spectral images[J]. Yellow River, 46(1): 93-96.
[34]	刘轩, 赵同谦, 蔡太义, 等, 2021. 丹江口水库总氮、氨氮遥感反演及时空变化研究[J]. 农业资源与环境学报, 38(5): 829-838.
	LIU X, ZHAO T Q, CAI T Y, et al., 2021. Spatiotemporal monitoring of total nitrogen and ammonia nitrogen in Danjiangkou reservoir[J]. Journal of Agricultural Resources and Environment, 38(5): 829-838.
[35]	王重洋, 周成虎, 陈水森, 等, 2021. 河口最大浑浊带研究的回顾与展望[J]. 科学通报, 66(18): 2328-2342.
	WANG C Y, ZHOU C H, CHEN S S, et al., 2021. Retrospect and perspective of the estuarine turbidity maximum zone researches[J]. Chinese Science Bulletin, 66(18): 2328-2342.
[36]	王思梦, 秦伯强, 2023. 湖泊水质参数遥感监测研究进展[J]. 环境科学, 44(3): 1228-1243.
	WANG S M, QIN B Q, 2023. Research progress on remote sensing monitoring of lake water quality parameters[J]. Environmental Science, 44(3): 1228-1243.

数据源及参考	波段	模型	r²	样本数	ρ/(mg·L^-1)	区域
Sentinel-2 (Dong et al., 2020)	B₃/B₂	y=0.296x²-0.224x-0.352	0.8510	70	0.004-0.355	丹江口水库
Sentinel-2 (Cao et al., 2023)	B₄/(B₅-B₁₂)	y=0.0585×e^-0.8332^x	0.7390	22	0.060-3.390	支脉河
Sentinel-2 (Tian et al., 2023)	B₂, B₃, B₄, B₅, B₆, B₇, B₈, B_8A	XGBoost	0.8200	96	0.030-0.810	清林径水库
Sentinel-2 (郭荣幸等, 2024)	B₈	y=30.545x²-13.034x+1.525	0.6135	18	0.110-0.350	小浪底水库
Sentinel-2 (Shi et al., 2024)	B₇/B₁	y=-2.633x³+9.269x²-10.16x+3.64	0.8024	120	0.025-0.400	淮河流域
Sentinel-2 (刘轩等, 2021)	B₃/B₂	BPNN	0.8770	70	0.001-0.067	丹江口水库
Sentinel-2 (Kuan et al., 2020)	B₂, B₃, B₄, B₅, B₈	BPNN	0.7400	41	0.100-0.980	淮河流域
UAV (Chen et al., 2021b)	(B₃+B₄)/B₂, (B₂+B₃+B₄)/B₂, (B₃+B₄)/(B₁+B₂), (B₂+B₃+B₄)/(B₃+B₄)	GA_XGBoost	0.6940	87	0.165-1.620	南淝河
UAV; GF-1C (Chen et al., 2023)	B₃/(B₁+B₄), B₄/B₃, (B₂+B₃+B₄)/ (B₁+B₄), (B₁+B₂+B₄)/(B₂+B₃), (B₁+B₂+B₄)/ (B₂+B₃), (B₂+B₃)/B₁+B₂+B₄), (B₂+B₃)/(B₁+B₄)	self-optimizing machine learning method	0.7790	77	0.433-1.396	南淝河
SPOT-5 (Wang et al., 2011)	B₁, B₂, B₃, B₄	GA-SVR	0.9824	13	0.150-0.500	渭河
Landsat 8 (Zhang et al., 2022)	B₁, B₂, B₃, B₄, B₅, B₈, B₉, B₂/B₃, B₂/B₄, B₄/B₃, B₂/B₅, (B₃-B₄)/(B₃+B₄), (B₅-B₄)/(B₅+B₄), (B₅-B₆)/(B₅+B₆), (B₆-B₇)/(B₆+B₇), (B₃-B₆)/(B₃+B₆)	ConvLSTM	0.8800	138	0.010-0.380	东平湖
Landsat 8 (Li et al., 2022)	B₁, B₂, B₃, B₄, B₅, B₇	ANN	0.4400	67	0.030-1.210	南渡河
Landsat 8 (Song et al., 2024)	B₁-B₅, B₄-B₅, B₆/B₇, B₃-B₅, (B₆-B₇)/(B₆+B₇), B₂-B₅, (B₃-B₅)/(B₃+B₅), B₅+B₆	RF	0.7084	40	0.220-0.550	呼伦湖

数据源及参考	波段	模型	r²	样本数	ρ/(mg·L^-1)	区域
Sentinel-2 (Dong et al., 2020)	B₃/B₂	y=0.296x²-0.224x-0.352	0.8510	70	0.004-0.355	丹江口水库
Sentinel-2 (Cao et al., 2023)	B₄/(B₅-B₁₂)	y=0.0585×e^-0.8332^x	0.7390	22	0.060-3.390	支脉河
Sentinel-2 (Tian et al., 2023)	B₂, B₃, B₄, B₅, B₆, B₇, B₈, B_8A	XGBoost	0.8200	96	0.030-0.810	清林径水库
Sentinel-2 (郭荣幸等, 2024)	B₈	y=30.545x²-13.034x+1.525	0.6135	18	0.110-0.350	小浪底水库
Sentinel-2 (Shi et al., 2024)	B₇/B₁	y=-2.633x³+9.269x²-10.16x+3.64	0.8024	120	0.025-0.400	淮河流域
Sentinel-2 (刘轩等, 2021)	B₃/B₂	BPNN	0.8770	70	0.001-0.067	丹江口水库
Sentinel-2 (Kuan et al., 2020)	B₂, B₃, B₄, B₅, B₈	BPNN	0.7400	41	0.100-0.980	淮河流域
UAV (Chen et al., 2021b)	(B₃+B₄)/B₂, (B₂+B₃+B₄)/B₂, (B₃+B₄)/(B₁+B₂), (B₂+B₃+B₄)/(B₃+B₄)	GA_XGBoost	0.6940	87	0.165-1.620	南淝河
UAV; GF-1C (Chen et al., 2023)	B₃/(B₁+B₄), B₄/B₃, (B₂+B₃+B₄)/ (B₁+B₄), (B₁+B₂+B₄)/(B₂+B₃), (B₁+B₂+B₄)/ (B₂+B₃), (B₂+B₃)/B₁+B₂+B₄), (B₂+B₃)/(B₁+B₄)	self-optimizing machine learning method	0.7790	77	0.433-1.396	南淝河
SPOT-5 (Wang et al., 2011)	B₁, B₂, B₃, B₄	GA-SVR	0.9824	13	0.150-0.500	渭河
Landsat 8 (Zhang et al., 2022)	B₁, B₂, B₃, B₄, B₅, B₈, B₉, B₂/B₃, B₂/B₄, B₄/B₃, B₂/B₅, (B₃-B₄)/(B₃+B₄), (B₅-B₄)/(B₅+B₄), (B₅-B₆)/(B₅+B₆), (B₆-B₇)/(B₆+B₇), (B₃-B₆)/(B₃+B₆)	ConvLSTM	0.8800	138	0.010-0.380	东平湖
Landsat 8 (Li et al., 2022)	B₁, B₂, B₃, B₄, B₅, B₇	ANN	0.4400	67	0.030-1.210	南渡河
Landsat 8 (Song et al., 2024)	B₁-B₅, B₄-B₅, B₆/B₇, B₃-B₅, (B₆-B₇)/(B₆+B₇), B₂-B₅, (B₃-B₅)/(B₃+B₅), B₅+B₆	RF	0.7084	40	0.220-0.550	呼伦湖

样本类型	Sentinel-2 影像日期	采样时间	样本数	NH₃-N质量浓度/(mg·L^-1)				变异系数
样本类型	Sentinel-2 影像日期	采样时间	样本数	最大值	最小值	平均值	标准差	变异系数
模型校准	3月11日 8月8日 11月6日 11月11日 12月6日 12月11日	3月11‒12日 8月7‒9日 11月5‒7日 11月11‒12日 12月5‒6日 12月10日	102	6.210	0.034	0.881	1.104	1.253
模型验证			44	2.560	0.026	0.645	0.648	1.004
独立验证			33	3.730	0.081	0.629	0.670	1.065
趋势检验	8月8日 9月22日 10月17日 11月6日 11月11日 12月6日	8月13‒14日 9月6‒10日 10月14‒17日 11月8‒13日 12月10日	25	2.86	0.137	1.135	0.650	0.573
总计	8景		204	6.210	0.026	0.821	0.922	1.124

样本类型	Sentinel-2 影像日期	采样时间	样本数	NH₃-N质量浓度/(mg·L^-1)				变异系数
样本类型	Sentinel-2 影像日期	采样时间	样本数	最大值	最小值	平均值	标准差	变异系数
模型校准	3月11日 8月8日 11月6日 11月11日 12月6日 12月11日	3月11‒12日 8月7‒9日 11月5‒7日 11月11‒12日 12月5‒6日 12月10日	102	6.210	0.034	0.881	1.104	1.253
模型验证			44	2.560	0.026	0.645	0.648	1.004
独立验证			33	3.730	0.081	0.629	0.670	1.065
趋势检验	8月8日 9月22日 10月17日 11月6日 11月11日 12月6日	8月13‒14日 9月6‒10日 10月14‒17日 11月8‒13日 12月10日	25	2.86	0.137	1.135	0.650	0.573
总计	8景		204	6.210	0.026	0.821	0.922	1.124

名称	波长/nm		空间分辨率/ m	描述
名称	S2A	S2B	空间分辨率/ m	描述
B₁	443.9	442.3	60	气溶胶
B₂	496.6	492.1	10	蓝色
B₃	560	559	10	绿色
B₄	664.5	665	10	红色
B₅	703.9	703.8	20	红边1
B₆	740.2	739.1	20	红边2
B₇	728.5	779.7	20	红边3
B₈	835.1	833	10	近红外
B_8A	864.8	864	20	红边4
B₉	945	943.2	60	水汽
B₁₀	1373.5	1376.9	60	卷云
B₁₁	1613.7	1610.4	20	短波红外1
B₁₂	2202.4	2185.7	20	短波红外2

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

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

图/表 17

参考文献 36

相关文章 6

编辑推荐

Metrics

本文评价

参照	反演模型	输入变量	r_c²	σ_RMSEc/(mg·L^-1)	σ_MAEc/(mg·L^-1)	r_v²	σ_RMSEv/(mg·L^-1)	σ_MAEv/(mg·L^-1)
Dong et al., 2020	y=0.25x²-0.62x+1.25	B₃/B₂	0.0003	1.104	0.799	-0.1256	0.688	0.594
郭荣幸等, 2024	y=27.37x²-8.11x+1.25	B₈	0.0164	1.095	0.775	-0.1960	0.709	0.618
Shi et al., 2024	y=0.05x³-0.45x²+0.95x +0.42	B₇/B₁	0.0235	1.091	0.783	-0.1516	0.696	0.613
Tian et al., 2023	XGBoost	B₂‒B_8A	0.4907	0.788	0.533	0.2146	0.575	0.456
刘轩等, 2021	BPNN	B₃/B₂	0.0001	1.105	0.802	0.0171	0.687	0.592
Kuan et al., 2020	BPNN	B₂‒B₅, B₈	0.1543	1.021	0.720	0.2473	0.584	0.491

指数	形式	指数	形式
S₁	B_i	S₁₈	(B_i-B_j)/B_k
S₂	B_i+B_j	S₁₉	(B_i×B_j)+B_k
S₃	B_i-B_j	S₂₀	(B_i×B_j)-B_k
S₄	B_i×B_j	S₂₁	B_i×B_j×B_k
S₅	B_i/B_j	S₂₂	B_i×B_j/B_k
S₆	(B_i+Bj)^0.5	S₂₃	B_i/B_j+B_k
S₇	(B_i/B_j)^0.5	S₂₄	B_i/B_j-B_k
S₈	(B_i×B_j)^0.5	S₂₅	B_i+1/B_j+1/B_k
S₉	(B_i²+B_j²)^0.5	S₂₆	B_i-1/B_j+1/B_k
S₁₀	(1/B_j+1/B_k)	S₂₇	B_i×(1/B_j+1/B_k)
S₁₁	(1/B_j-1/B_k)	S₂₈	B_i/(1/B_j+1/B_k)
S₁₂	B_i+B_j+B_k	S₂₉	B_i+1/B_j-1/B_k
S₁₃	B_i+B_j-B_k	S₃₀	B_i-1/B_j-1/B_k
S₁₄	(B_i+B_j)×B_k	S₃₁	B_i×(1/B_j-1/B_k)
S₁₅	(B_i+B_j)/B_k	S₃₂	B_i/(1/B_j-1/B_k)
S₁₆	B_i-B_j-B_k	S₃₃	(B_i-B_j)/(B_i+B_j)
S₁₇	(B_i-B_j)×B_k	S₃₄	(B_i+B_j)/(B_k+B_m)

模型	r_c²	σ_RMSEc/ (mg·L^-1)	σ_MAEc/ (mg·L^-1)	r_v²	σ_RMSEv/ (mg·L^-1)	σ_MAEv/ (mg·L^-1)
BC-FDR_XGBoost	0.6872	0.617	0.385	0.5436	0.438	0.362
BC-FDR_RF	0.5624	0.854	0.497	0.4237	0.494	0.414
BC-FDR_SVR	0.1935	0.992	0.639	0.1795	0.590	0.479

Sentinel-2反演结果	丰水期	枯水期
平均值/(mg·L^-1)	0.552	0.795
质量浓度区间/(mg·L^-1)	0.182‒2.422	0.153‒2.510

[1]	王薇, 夏宇轩. 基于遥感技术和机器学习的城市街区PM_2.5空间分布特征研究——以合肥市滨湖新区为例[J]. 生态环境学报, 2024, 33(9): 1426-1437.
[2]	杨可明, 彭里顺, 张燕海, 谷新茹, 陈新阳, 江克贵. 淮北矿区多种类型植被地上生物量反演研究[J]. 生态环境学报, 2024, 33(7): 1027-1035.
[3]	王雪梅, 杨雪峰, 赵枫, 安柏耸, 黄晓宇. 基于机器学习算法的干旱区绿洲地上生物量估算[J]. 生态环境学报, 2023, 32(6): 1007-1015.
[4]	姜鹏, 秦美欧, 李荣平, 孟莹, 杨霏云, 温日红, 孙沛, 方缘. 中国典型生态系统GPP的季节变异及其影响要素[J]. 生态环境学报, 2022, 31(4): 643-651.
[5]	谢邵文, 郭晓淞, 杨芬, 黄强, 陈曼佳, 魏兴琥, 刘承帅. 广州市城市公园土壤重金属累积特征、形态分布及其生态风险[J]. 生态环境学报, 2022, 31(11): 2206-2215.
[6]	陈漾, 张金谱, 邱晓暖, 琚鸿, 黄俊. 2021年广州市臭氧污染特征及气象因子影响分析[J]. 生态环境学报, 2022, 31(10): 2028-2038.