Factors affecting the ability of extensive green roofs to reduce nutrient pollutants in rainfall runoff Yongwei Gong a , Xianwei Zhang a , Junqi Li a, ⁎, Xing Fang b , Dingkun Yin a,c , Peng Xie a , Linmei Nie d a Key Laboratory of Urban Stormwater System and Water Environment, Ministry of Education, Beijing University of Civil Engineering and Architecture, Beijing 100044, China b Department of Civil Engineering, Auburn University, Auburn, AL 36849-5337, USA c School of Environment, Tsinghua University, Beijing 100084, China d Centre for Sustainable Development and Innovation of Water Technology, Oslo 0957, Norway HIGHLIGHTS • Factors affecting EGR's ability to reduce nutrient pollution were analysed. • Increasing EGR module scale increased nitrogen load removal rate. • Growing medium materials signifi- cantly influenced phosphorus removal. • Rainfall depth significantly influenced nutrient reduction. GRAPHICAL ABSTRACT abstract article info Article history: Received 26 March 2020 Received in revised form 26 April 2020 Accepted 4 May 2020 Available online xxxx Editor: José Virgílio Cruz Keywords: Extensive green roofs Nutrient concentration Nutrient load reduction Structural materials Influential factors Green roofs can retain urban rainfall runoff, but there are doubts about whether they can reduce urban nonpoint source pollution. To explore the factors affecting the ability of green roofs to reduce nutrients in rainfall runoff, nine types of extensive green roofs (EGRs) were analysed during 38 natural rainfall events and two early spring irrigation runoff events from 1 March to 30 November 2019 in Beijing. Differences among the module scale, growing medium material, growing medium depth, drainage layer material, planting time, rainfall characteristics and seasonal variation were examined to study their correlation with pollutant event mean concentration (EMC) and the load reduction performance of EGRs. The results showed that EGRs had higher total nitrogen (TN), am- monia nitrogen (NH 4 + -N) and nitrate nitrogen (NO 3 - -N) concentrations than traditional concrete roofs, but total phosphorus (TP) concentrations were similar, and EGRs could reduce some of the nutrient loads. One-way anal- ysis of variance showed that the module scale, growing medium material, growing medium depth, drainage layer material, and planting time had no significant effect on TN and NO 3 - -N concentrations (p N 0.05). The growing medium material had a significant effect on the TP concentration (p b 0.05). From the perspective of nutrient load reduction, module scale had a significant effect on TN and NH 4 + -N loads (p b 0.05). The growing medium depth had a significant effect on NH 4 + -N loads (p b 0.05). In addition, the growing medium material had a signif- icant effect on TP loads (p b 0.05). When porous wool fibre and a bumpy plastic drainage board were selected as drainage layer materials, the effect on the NO 3 - -N load differed significantly. In other situations, there were no Science of the Total Environment 732 (2020) 139248 ⁎ Corresponding author. E-mail address: lijunqi@bucea.edu.cn (J. Li). https://doi.org/10.1016/j.scitotenv.2020.139248 0048-9697/© 2020 Elsevier B.V. All rights reserved. Contents lists available at ScienceDirect Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv