Experimental and numerical investigation on hydrological characteristics of extensive green roofs under the influence of rainstorms

Green roof rainwater retention, peak runoff reduction, and runoff time delay are considered important hydrological performance indicators for assessing management of urban stormwater. In this study, simulated rainfall experiments were conducted on three green roof models with different water storage layer depths. The numerical model was established using Hydrus-1D program, and the sensitivity of main parameters, the hydrological response of green roofs with a water storage layer, and water storage on the soil surface were analyzed. In addition to the saturated water content of the soil, the depth of the green roof water storage layer is the most sensitive parameter to rainwater retention and initial drainage time. During the simulated rainfall experiment, the 25-mm-deep water storage layer (WSL-25) increased the rainwater retention capacity (RRC) by 46%. For a 20-year return period corresponding to South China region, the RRC of green roofs with WSL-25 increased by 31% compared with that without a water storage layer. The initial drainage time was delayed by 50 min, and the peak drainage rate was reduced by 89%. In this case, a 100-mm soil layer, a 50-mm water storage layer, and a 50 mm maximum surface water storage depth were considered the optimal structural configurations of green roofs. This shows that water storage on the soil surface and bottom water storage layer were equally important for improving RRC, reducing peak drainage and delaying drainage time of green roofs.