[Integrated Assessment of Runoff Quality from Green Roofs with Different Configurations].

Green roofs are regarded as one of the important measures for the sponge city construction. However, the comprehensive impacts of configuration factors (e.g., vegetation and substrates) on runoff quality from green roofs are not clear, which limits the promotion of green roofs. In this study, 12 green roofs with three vegetation types (i.e., Sedum lineare, Portulaca grandiflora, and non-vegetated substrates), three substrate types (i.e., local planting soil, engineered soil, and light growing medium), and two substrate depths (i.e., 10 cm and 15 cm) were set up in Beijing. During the rainy season of 2019, the rainfall characteristics, runoff volumes, and concentrations of nutrients and heavy metals of runoff from the green roofs were monitored. Based on the measured data, a runoff quality index (RQI) was developed to evaluate the comprehensive influences of configurations on runoff quality of the green roofs. The results showed that vegetation could improve runoff reduction rate and decrease the concentrations of NO3--N in runoff of green roofs. The RQIs of green roofs planted with S. linear and P. grandiflora were similar, and the evaluation results of runoff quality were better than those with non-vegetated substrates. The materials of substrates had significant influences on the runoff reduction rate and pollutant concentrations in runoff from green roofs. The green roofs with light growing medium, which had the lowest runoff reduction rates and the highest concentrations of NH4+-N, DFe, DMn, and DZn in the runoff, showed poorer runoff quality than the green roofs with local planting soil and engineered soil. The green roofs with a substrate depth of 15 cm had higher runoff reduction rates than those with 10 cm deep substrate, and the runoff quality was better than those with a substrate depth of 10 cm. The results of this study provide scientific reference for the design and integrated assessment of green roofs.

[Inter-annual Changes in Runoff Quality from Green Roofs with Different Vegetation].

As an important measure of the sponge city, green roofs have received extensive attention in recent years. To investigate the inter-annual changes in runoff quality of green roofs with different vegetation types, three green roofs with different vegetation cover (Sedum lineare, Portulaca grandiflora, and a non-vegetated control) were set up in Beijing. The influences of vegetation and monitoring period on runoff quality from the green roofs were evaluated using the plant growth characteristics and the quality of rainwater and runoff from the green roofs during the rainy season of 2017-2019. The results showed that all three green roofs were the sinks of NH4+-N, and the average mass concentration reduction rates were between 50.1% and 79.2%. However, all three green roofs were sources of PO43--P, DCr, DCu, and DNi. The green roofs covered with S. lineare and P. grandiflora were sinks of NO3--N in 2017, and the average mass concentration reduction rates were 71.4% and 99.5%, respectively, but they became sources of NO3--N in both 2018 and 2019. However, the non-vegetated control was the source of NO3--N in all three rainy seasons. Both vegetation type and length of monitoring period had significant effects on the mass concentrations of NO3--N, PO43--P, DNi, and DCu in runoff from the green roofs (P<0.05) but had no significant effects on the mass concentrations of NH4+-N and DCr in runoff from the green roofs (P>0.05). In 2017-2019, the mass concentrations of NO3--N in runoff from the non-vegetated control and the green roofs covered by S. lineare and the mass concentration of PO43--P in runoff from the green roof covered by P. grandiflora increased yearly. The mass concentrations of DNi and DCu in runoff from all three green roofs increased in 2018 but dropped in 2019. Among the green roofs with different vegetation types, the green roof covered by P. grandiflora showed better NO3--N retention capacity than that of the other green roofs but may have increased the concentrations of PO43--P, DNi, and DCu in the runoff.