Predicting and Managing Water Quality in Complex River Networks Using a Machine Learning Approach: A Case Study of the Qinhuai River Basin
DOI:
https://doi.org/10.15377/2410-3624.2026.13.5Keywords:
Machine learning, Scenario simulation, Water quality prediction, River basin management, Non-point source pollution.Abstract
As water pollution intensifies, traditional monitoring and modeling methods struggle to address the complexities of dynamic water environments, particularly in complex river networks. Machine learning, a powerful data-driven model, has become a key research focus in water resource management due to its accurate predictive capabilities. This study examines the Qinhuai River Basin (QRB) in the lower Yangtze River, using water quality data from 10 monitoring stations (2021-2023) for dissolved oxygen (DO), permanganate index (CODMn), ammonia nitrogen (NH₃-N), and total phosphorus (TP). The study first analyzes the spatiotemporal distribution of water quality, then develops machine learning models to identify key factors affecting water quality at monitoring points. Finally, scenario simulations and management strategies are proposed. The results show that upstream DO and CODMn concentrations are higher, while NH₃-N and TP are lower compared to downstream. Summer is the primary period for water quality exceedance, with DO as the key indicator. XGBoost outperforms other models with R2 values of 0.882–0.947 for CODMn and 0.824–0.890 for NH₃-N. Key factors influencing water quality include upstream water quality, 48-hour cumulative rainfall runoff, basin inflow and outflow, and wastewater treatment plant (WWTP) effluent. Scenario simulations reveal that improving upstream water quality to Class III standards leads to downstream improvements, while rainfall runoff interception, ecological water replenishment, and reducing WWTP discharge can effectively reduce water quality indicator concentrations. This study provides useful insights into water quality management and demonstrates the potential of the proposed framework as a transferable tool for supporting adaptive water quality management in other data-limited and complex river basins.
References
[1] Vega M, Pardo R, Barrado E, Debán L. Assessment of seasonal and polluting effects on the quality of river water by exploratory data analysis. Water Res. 1998; 32(12): 3581-92. https://doi.org/10.1016/S0043-1354(98)00138-9
[2] Grill G, Lehner B, Thieme M, Geenen B, Tickner D, Antonelli F, et al. Mapping the world's free-flowing rivers. Nature. 2019; 569(7755): 215-21. https://doi.org/10.1038/s41586-019-1111-9
[3] Palmer M, Ruhi A. Linkages between flow regime, biota, and ecosystem processes: Implications for river restoration. Science. 2019; 365(6459): eaaw2087. https://doi.org/10.1126/science.aaw2087
[4] Chen SK, Jang CS, Chou CY. Assessment of spatiotemporal variations in river water quality for sustainable environmental and recreational management in the highly urbanized Danshui River basin. Environ Monit Assess. 2019; 191(2): 100. https://doi.org/10.1007/s10661-019-7246-1
[5] Strokal M, Ma L, Bai Z, Luan S, Kroeze C, Oenema O, et al. Alarming nutrient pollution of Chinese rivers as a result of agricultural transitions. Environ Res Lett. 2016; 11(2): 024014. https://doi.org/10.1088/1748-9326/11/2/024014
[6] Xiao Q, Hu Z, Hu C, Islam ARMT, Bian H, Chen S, et al. A highly agricultural river network in Jurong Reservoir watershed as significant CO₂ and CH₄ sources. Sci Total Environ. 2021; 769: 144558. https://doi.org/10.1016/j.scitotenv.2020.144558
[7] Kim J, Lee T, Seo D. Algal bloom prediction of the lower Han River, Korea using the EFDC hydrodynamic and water quality model. Ecol Model. 2017; 366: 27-36. https://doi.org/10.1016/j.ecolmodel.2017.10.015
[8] Shin S, Her Y, Muñoz-Carpena R, Yu X, Martinez C, Singh A. Climate change impacts on water quantity and quality of a watershed-lake system using a spatially integrated modeling framework in the Kissimmee River-Lake Okeechobee system. J Hydrol Reg Stud. 2023; 47: 101408. https://doi.org/10.1016/j.ejrh.2023.101408
[9] Yu J, Tian Y, Wang X, Sun T, Lancia M, Andrews CB, et al. Integrated modeling of flow, soil erosion, and nutrient dynamics in a regional watershed: Assessing natural and human-induced impacts. Water Resour Res. 2024; 60(9): e2024WR037531. https://doi.org/10.1029/2024WR037531
[10] Peng K, Li J, Zhou X, Li H, Xie W, Zhang K, et al. Simulation and control of non-point source pollution based on MIKE model: A case study of Danjiang River Basin, China. Ecohydrol Hydrobiol. 2023; 23(4): 554-68. https://doi.org/10.1016/j.ecohyd.2023.08.001
[11] Jiang L, Li Y, Zhao X, Tillotson MR, Wang W, Zhang S, et al. Parameter uncertainty and sensitivity analysis of water quality model in Lake Taihu, China. Ecol Model. 2018; 375: 1-12. https://doi.org/10.1016/j.ecolmodel.2018.02.014
[12] Li S, Feng J, Zhang S. Simulation analysis of water quality improvement effect of ecological water replenishment measures in Qilu Lake (In Chinese). Water Resour Power. 2020; 38(11): 35-9.
[13] Zanoni MG, Majone B, Bellin A. A catchment-scale model of river water quality by machine learning. Sci Total Environ. 2022; 838: 156377. https://doi.org/10.1016/j.scitotenv.2022.156377
[14] Gao Z, Chen J, Wang G, Ren S, Fang L, Yinglan A, et al. A novel multivariate time series prediction of crucial water quality parameters with long short-term memory (LSTM) networks. J Contam Hydrol. 2023; 259: 104262. https://doi.org/10.1016/j.jconhyd.2023.104262
[15] Liang Z, Zou R, Chen X, Ren T, Su H, Liu Y. Simulate the forecast capacity of a complicated water quality model using the long short-term memory approach. J Hydrol. 2020; 581: 124432. https://doi.org/10.1016/j.jhydrol.2019.124432
[16] Hong SM, Abbas A, Kim S, Kwon DH, Yoon N, Yun D, et al. Autonomous calibration of EFDC for predicting chlorophyll-a using reinforcement learning and a real-time monitoring system. Environ Model Softw. 2023; 168: 105805. https://doi.org/10.1016/j.envsoft.2023.105805
[17] Keller AA, Garner K, Rao N, Knipping E, Thomas J. Hydrological models for climate-based assessments at the watershed scale: A critical review of existing hydrologic and water quality models. Sci Total Environ. 2023; 867: 161209. https://doi.org/10.1016/j.scitotenv.2022.161209
[18] Zhong S, Zhang K, Bagheri M, Burken JG, Gu A, Li B, et al. Machine learning: New ideas and tools in environmental science and engineering. Environ Sci Technol. 2021; 55(19): 12741-54. https://doi.org/10.1021/acs.est.1c01339
[19] Chen K, Chen H, Zhou C, Huang Y, Qi X, Shen R, et al. Comparative analysis of surface water quality prediction performance and identification of key water parameters using different machine learning models based on big data. Water Res. 2020; 171: 115454. https://doi.org/10.1016/j.watres.2019.115454
[20] Kumar N, Sinha A, Pasupuleti S. Application of machine learning models for water quality prediction in a coal-mining region: Insights into fluoride and phenol contamination in Dhanbad District, Eastern India. Environ Pollut. 2026; 405: 128558. https://doi.org/10.1016/j.envpol.2026.128558
[21] Tong X, Zhang X, Xiao S, Chen J, Zhou C, Zhou J, et al. A machine learning-assisted integrated framework for linking water quality and phytoplankton carbon fixation potential in drinking water reservoirs. Water Res. 2026; 302: 126104. https://doi.org/10.1016/j.watres.2026.126104
[22] Yu L, Liu X, Hua Z. Occurrence, distribution, and risk assessment of perfluoroalkyl acids in drinking water sources from the lower Yangtze River. Chemosphere. 2022; 287: 132064. https://doi.org/10.1016/j.chemosphere.2021.132064
[23] Xu Z. Analysis of key elements in water environment governance of Yangtze River (In Chinese). Water Resour Dev Res. 1-13.
[24] Ma Y, Chen X, Zhang B. Screening of characteristic pollutants and risk pollutants in different media in the Yangtze River Basin using comprehensive evaluation and risk quotient methods (In Chinese). Environ Sci. 2024; 45(9): 5214-26.
[25] Zou H, Li J, Cai Y, Wang X, Ma T, Duan X. Characterizing the spatial variations of the relationship between chemical industry and riparian water quality along the Yangtze River. Appl Geogr. 2026; 194: 104117. https://doi.org/10.1016/j.apgeog.2026.104117
[26] Zhang C, Li K, Fang Q, Zhang S. Construction and application of an aquatic ecological health assessment system based on multi-source data and multi-method integration: A case study of key water bodies in the Yangtze River. Ecol Indic. 2026; 188: 115038. https://doi.org/10.1016/j.ecolind.2026.115038
[27] Wang S, Pu Y, Li S, Li R, Li M. Spatio-temporal analysis of impervious surface expansion in the Qinhuai River Basin, China, 1988-2017. Remote Sens (Basel). 2021; 13(22): 4494. https://doi.org/10.3390/rs13224494
[28] Du J, Qian L, Rui H, Zuo T, Zheng D, Xu Y, et al. Assessing the effects of urbanization on annual runoff and flood events using an integrated hydrological modeling system for Qinhuai River Basin, China. J Hydrol. 2012; 464-465: 127-39. https://doi.org/10.1016/j.jhydrol.2012.06.057
[29] Li J, Li Y, Qian B, Niu L, Zhang W, Cai W, et al. Development and validation of a bacteria-based index of biotic integrity for assessing the ecological status of urban rivers: A case study of Qinhuai River Basin in Nanjing, China. J Environ Manage. 2017; 196: 161-7. https://doi.org/10.1016/j.jenvman.2017.03.003
[30] Gao Y, Chen J, Luo H, Wang H. Prediction of hydrological responses to land use change. Sci Total Environ. 2020; 708: 134998. https://doi.org/10.1016/j.scitotenv.2019.134998
[31] Liu M, Wang Y, Liu Y. Analysis of uneven distribution of precipitation during flood season in the Qinhuai River Basin (In Chinese). Water Resour Dev Manag. 2024; 10(5): 65-9.
[32] Bian G, Wang G, Chen J, Zhang J, Song M. Spatial and seasonal variations of hydrological responses to climate and land-use changes in a highly urbanized basin of Southeastern China. Hydrol Res. 2021; 52(2): 506-22. https://doi.org/10.2166/nh.2021.087
[33] Li R, Du J, Bian G, Wang Y, Chen C, Zhang X, et al. An integrated modelling approach for flood simulation in the urbanized Qinhuai River Basin, China. Water Resour Manag. 2020; 34(13): 3967-84. https://doi.org/10.1007/s11269-020-02614-4
[34] Abro MI, Zhu D, Wei M, Majidano AA, Khaskheli MA, Abideen ZU, et al. Hydrological appraisal of rainfall estimates from radar, satellite, raingauge and satellite-gauge combination on the Qinhuai River Basin, China. Hydrol Sci J. 2019; 64(16): 1957-71. https://doi.org/10.1080/02626667.2018.1557335
[35] Chen R, Hu Q, Shen W, Guo J, Yang L, Yuan Q, et al. Identification of nitrate sources of groundwater and rivers in complex urban environments based on isotopic and hydro-chemical evidence. Sci Total Environ. 2023; 871: 162026. https://doi.org/10.1016/j.scitotenv.2023.162026
[36] Xia Y, Ti C, She D, Yan X. Linking river nutrient concentrations to land use and rainfall in a paddy agriculture-urban area gradient watershed in southeast China. Sci Total Environ. 2016; 566-567: 1094-105. https://doi.org/10.1016/j.scitotenv.2016.05.134
[37] Yan X, Han H, Qiu J, Zhang L, Xia Y, Yan X. Suburban agriculture increased N levels but decreased indirect N₂O emissions in an agricultural-urban gradient river. Water Res. 2022; 220: 118639. https://doi.org/10.1016/j.watres.2022.118639
[38] Sang C, Wang J, Guo Y, Xie C, Zhou Q. Integrated evaluation of water quality in semi-arid Eurasian rivers using the methods of single-factor and water quality index. Ecol Indic. 2026; 183: 114674. https://doi.org/10.1016/j.ecolind.2026.114674
[39] Su K, Wang Q, Li L, Cao R, Xi Y. Water quality assessment of Lugu Lake based on Nemerow pollution index method. Sci Rep. 2022; 12(1): 13613. https://doi.org/10.1038/s41598-022-17874-w
[40] State Environmental Protection Administration. Environmental quality standards for surface water (PR China) (GB 3838-2002) [S]. Beijing: China Environmental Science Press; 2002.
[41] Wang Q, Li Z, Cai J, Zhang M, Liu Z, Xu Y, et al. Spatially adaptive machine learning models for predicting water quality in Hong Kong. J Hydrol. 2023; 622: 129649. https://doi.org/10.1016/j.jhydrol.2023.129649
[42] Zhang Q, Li Z, Zhu L, Zhang F, Sekerinski E, Han JC, et al. Real-time prediction of river chloride concentration using ensemble learning. Environ Pollut. 2021; 291: 118116. https://doi.org/10.1016/j.envpol.2021.118116
[43] Qiu R, Wang Y, Rhoads B, Wang D, Qiu W, Tao Y, et al. River water temperature forecasting using a deep learning method. J Hydrol. 2021; 595: 126016. https://doi.org/10.1016/j.jhydrol.2021.126016
[44] Ai H, Zhang K, Sun J, Zhang H. Short-term Lake Erie algal bloom prediction by classification and regression models. Water Res. 2023; 232: 119710. https://doi.org/10.1016/j.watres.2023.119710
[45] Wang R, Kim JH, Li MH. Predicting stream water quality under different urban development pattern scenarios with an interpretable machine learning approach. Sci Total Environ. 2021; 761: 144057. https://doi.org/10.1016/j.scitotenv.2020.144057
[46] Li Z, Gu H, Kang Y, Zhang C, Niu Z, Shen D. Spatiotemporal assessment of water damage susceptibility in China's road infrastructure: A machine learning and SHAP approach using social media data. J Hydrol. 2026; 664: 134539. https://doi.org/10.1016/j.jhydrol.2025.134539
[47] Friedman JH. Greedy function approximation: A gradient boosting machine. Ann Stat. 2001; 29: 1189-232. https://doi.org/10.1214/aos/1013203451
Downloads
Published
Issue
Section
License
Copyright (c) 2026 Sheng Zhang, Shaoyan Feng, Xin Qian
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
All the published articles are licensed under the terms of the Creative Commons Attribution Non-Commercial License (CC BY-NC 4.0) which permits unrestricted, non-commercial use, distribution and reproduction in any medium, provided the work is properly cited.
