PDF Download

[1]Bi Jieyu,Li Ruixian,Zhang Shouhong,et al.Evaluation and simulation of roof rainwater runoff quality in Beijing based on explainable machine learning[J].Research of Soil and Water Conservation,2025,32(06):280-289.[doi:10.13869/j.cnki.rswc.2025.06.020]

Copy

Evaluation and simulation of roof rainwater runoff quality in Beijing based on explainable machine learning

Research of Soil and Water Conservation[ISSN 1005-3409/CN 61-1272/P] Volume: 32 Number of periods: 2025 06 Page number: 280-289 Column: Public date: 2025-10-20

Title:: Evaluation and simulation of roof rainwater runoff quality in Beijing based on explainable machine learning

Author(s):: Bi Jieyu¹,Li Ruixian¹,Zhang Shouhong^1,2,3,Zhang Fan¹,Zhang Sunxun⁴,Wei Zhen⁵; (1.School of Soil and Water Conservation,Beijing Forestry University,Beijing 100083,China;2.JixianNational Forest Ecosystem Observation and Research Station,Linfen,Shanxi 042200,China;3.Beijing Engineering Research Center of Soil and Water Conservation,Beijing 100083,China;4.State Key Laboratory of Simulation and Regulation of Water Cycle in River Basin,China Institute of Water Resources and Hydropower Research,Beijing 100048,China;5.Beijing Songshan National Nature Reserve...)

Keywords:: rainwater harvesting system; runoff quality; machine learning; SHAP model

CLC:: TU991.34

DOI:: 10.13869/j.cnki.rswc.2025.06.020

Abstract:: [Objective] This study aims to quantitatively evaluate the runoff quality characteristics of roof rainwater harvesting system under differentiated meteorological and environmental factors in Beijing, and to analyze their primary influencing factors, to provide a scientific basis for the design of urban roof rainwater harvesting systems and comprehensive evaluation of runoff quality. [Methods] From July to November 2023, experimental rainwater harvesting system was installed at five sites across different districts of Beijing to monitor environmental factors, rainfall characteristics, and runoff quality data. The Runoff Quality Index(RQI) was used for runoff quality evaluation, and an explainable random forest regression model was established to predict RQI based on meteorological and environmental factors. [Results] (1) Pollutant concentrations in runoff exhibited distinct distribution patterns across regions. Haidian and Yanqing districts showed higher RQI values, while Fangshan had the lowest.(2) The random forest model achieved relatively accurate prediction of RQI values, demonstrating high precision(R2 =0.853 for calibration, R2 =0.745 for validation).(3) SHAP model analysis indicated that rainfall had the greatest influence on RQI, with a critical threshold of approximately 8 mm for the first-flush effect. [Conclusion] Significant differences in roof runoff quality were observed across different districts of Beijing, primarily driven by rainfall, antecedent dry period, and atmospheric NO2 concentration. The random forest model effectively predicted runoff quality, and its integration with the SHAP model enabled quantitative assessment of the contribution of each influencing factor.

References:: [1]Xu J Y，Dai J Y，Wu X F，et al.Urban rainwater utilization： a review of management modes and harvesting systems[J]. Frontiers in Environmental Science，2023，11：1025665.
[2]Teston A，Piccinini Scolaro T，Kuntz Maykot J，et al. Comprehensive environmental assessment of rainwater harvesting systems：a literature review[J]. Water，2022，14（17）：2716.
[3]Zhang Q Q，Wang X K，Hou P Q，et al.Quality and seasonal variation of rainwater harvested from concrete，asphalt，ceramic tile and green roofs in Chongqing，China [J]. Journal of Environmental Management，2014，132： 178-187.
[4]Alim M A，Rahman A，Tao Z，et al.Suitability of roof harvested rainwater for potential potable water production：a scoping review[J]. Journal of Cleaner Production，2020，248：119226.
[5]Mao J，Xia B Y，Zhou Y，et al.Effect of roof materials and weather patterns on the quality of harvested rainwater in Shanghai，China[J]. Journal of Cleaner Production，2021，279：123419.
[6]Vieira-Filho M S，Lehmann C，Fornaro A.Influence of local sources and topography on air quality and rainwater composition in Cubat?o and S?o Paulo，Brazil[J]. Atmospheric Environment，2015，101：200-208.
[7]Murphy L U，O’Sullivan A，Cochrane T A.Quantifying the spatial variability of airborne pollutants to storm water runoff in different land-use catchments[J]. Water，Air，& Soil Pollution，2014，225（7）：2016.
[8]An M Y，Han Y G，Zhao C Y，et al.Effects of different wind directions on soil erosion and nitrogen loss processes under simulated wind-driven rain[J]. Catena，2022，217：106423.
[9]Ren M F，Xu Z X，Pang B，et al.Spatiotemporal variability of precipitation in Beijing，China during the wet seasons[J]. Water，2020，12（3）：716.
[10]北京市生态环境局.北京市环境状况公报[R]. 北京：北京市生态环境局，2023. Beijing Municipal Ecology and Environment Bureau. Beijing environmental status bulletin[R]. Beijing：Beijing Municipal Ecology and Environment Bureau，2023.
[11]曹建业.地表水水质评价与预警预测模型研究：以东莞市为例[D]. 广州：华南理工大学，2023. Cao J Y.Surface Water Quality Assessment and Early Warning Prediction Models：A Case Study of Dongguan City [D]. Guangzhou：South China University of Technology，2023.
[12]Yan T，Zhou A N，Shen S L.Prediction of long-term water quality using machine learning enhanced by Bayesian optimisation[J]. Environmental Pollution，2023，318：120870.
[13]Zhu M Y，Wang J W，Yang X，et al.A review of the application of machine learning in water quality evaluation[J]. Eco-Environment & Health，2022，1（2）： 107-116.
[14]张冲，陈末，刘袁媛，等.基于机器学习的气象水质耦合预测模型研究[J]. 水文，2024，44（6）：9-14，21. Zhang C，Chen M，Liu Y Y，et al.Research on coupled prediction model of meteorological and water quality based on machine learning[J]. Journal of China Hydrology，2024，44（6）：9-14，21.
[15]潘梦瑶，任瑛，王思源，等.基于梯度提升算法和SHAP的石家庄PM2.5和臭氧浓度预测及影响因素分析[J]. 环境科学学报，2024，44（7）：402-409. Pan M Y，Ren Y，Wang S Y，et al.Prediction of PM2.5 and ozone concentration in Shijiazhuang and analysis of influencing factors based on gradient boosting algorithm and SHAP[J]. Acta Scientiae Circumstantiae，2024，44 （7）：402-409.
[16]北京市水务局.北京市水资源公报[R]. 北京：北京市水务局，2023. Beijing Municipal Water Authority.Beijing water resources bulletin[R]. Beijing：Beijing Municipal Water Authority，2023.
[17]Fan C Z，Liu Y B，Liu C H，et al.Water quality characteristics，sources，and assessment of surface water in an industrial mining city，southwest of China[J]. Environmental Monitoring and Assessment，2022，194 （4）：259.
[18]严彦文，谢碧霄，孟得新，等.北京市空气质量的时间特征及影响因素分析[J]. 应用数学进展，2024，13（1）：118-126. Yan Y W，Xie B X，Meng D X，et al.Analysis of temporal characteristics and influencing factors of air quality in Beijing [J]. Advances in Applied Mathematics，2024，13（1）：118-126.
[19]Zhang W，Li J，Sun H C，et al.Pollutant first flush identification and its implications for urban runoff pollution control：a roof and road runoff case study in Beijing，China[J]. Water Science and Technology，2021，83（11）：2829-2840.
[20]Viancelli A，Michelon W.Climate change and nitrogen dynamics：challenges and strategies for a sustainable future[J]. Nitrogen，2024，5（3）：688-701.
[21]杨文琴，刘思慧，苗淼，等.北京市内到郊区氮沉降时空变化特征[J]. 环境科学学报，2016，36（5）：1530-1538. Yang W Q，Liu S H，Miao M，et al.Spatial and temporal variation of dissolved inorganic nitrogen deposition along an urban to rural transect in Beijing[J]. Acta Scientiae Circumstantiae，2016，36（5）：1530-1538.
[22]Shajib M T I，Hansen H C B，Liang T，et al.Metals in surface specific urban runoff in Beijing[J]. Environmental Pollution，2019，248：584-598.
[23]Sanjeeva A，Puttaswamaiah S G.Influence of atmospheric deposition and roof materials on harvested rainwater quality[J]. Journal of Environmental Engineering，2018，144（12）：04018121.
[24]Uchiyama R，Okochi H，Katsumi N，et al.The impact of air pollutants on rainwater chemistry during “urbaninduced heavy rainfall” in downtown Tokyo，Japan[J]. Journal of Geophysical Research：Atmospheres，2017，122（12）：6502-6519.
[25]Ding J，Jiang Y，Liu Q，et al.Influences of the land use pattern on water quality in low-order streams of the Dongjiang River basin，China：A multi-scale analysis[J]. Science of the Total Environment，2016，551：205-216.
[26]Wu J，Zeng S D，Yang L H，et al.Spatiotemporal characteristics of the water quality and its multiscale relationship with land use in the Yangtze River basin[J]. Remote Sensing，2021，13（16）：3309.
[27]Nosrati K.Identification of a water quality indicator for urban roof runoff[J]. Sustainability of Water Quality and Ecology，2017，9：78-87.
[28]Friedler E，Gilboa Y，Muklada H.Quality of roof-harvested rainwater as a function of environmental and air pollution factors in a coastal Mediterranean city（Haifa，Israel）[J]. Water，2017，9（11）：896.
[29]Kassomenos P A，Vardoulakis S，Chaloulakou A，et al.Study of PM10 and PM2.5 levels in three European cities：analysis of intra and inter urban variations[J]. Atmospheric Environment，2014，87：153-163.
[30]Sánchez A S，Cohim E，Kalid R A.A review on physicochemical and microbiological contamination of roofharvested rainwater in urban areas[J]. Sustainability of Water Quality and Ecology，2015，6：119-137.

Similar References:

Memo

Last Update: 2025-12-10