Abstract: Global climate change and the expansion of impervious surface in urban areas have increased the risk of urban flood. As a measure with low impact development (LID), roof greening can effectively reduce stormwater runoff, serving as the first barrier to prevent rainwater from entering the urban drainage system. We used the CITYgreen model to simulate and analyze the impacts of roof greening on hydrological parameters (e.g., surface runoff) in the new and old residential areas and the commercial areas in Nanjing City, and further investigated the variations in stormwater runoff effects (SRE) among these functional areas. We compared the SRE between different green roof types, and between green roofs and ground-level green spaces. The results showed that the proportion of permeable surface would increase by 28.9%, 12.5% and 49.2% respectively in the old residential area, new residential area and commercial area if all building roofs were greened. In a rainfall event with a return period of two years and a duration of 24 hours (precipitation=72 mm), the implementation of roof greening in all buildings in the three sampling areas could reduce surface runoff by 0-19.8% and reduce peak flow by 0-26.5%. The reductions in runoff by green roofs could be translated to a rainwater storage capacity of 223-2299 m³. The commercial area had the highest SRE by installing green roofs, followed by the old residential area, while the new residential area had the lowest SRE. The rainwater storage volume by per unit area of extensive green roof was about 78.6%-91.7% that of the intensive green roof. The storage capacity by per unit area of green roof was 31%-43% of that in the ground-level greenery. The results would provide scientific references for the site selection, sustainable design and incentive development of roof greening from the perspective of stormwater management.

Key words: climate change, low impact development, CITYgreen model, stormwater runoff effect