Abstract This paper investigates the performance of modular-tray green roof systems on rainwater retention and total runoff delay under heavy rain conditions, typical of countries with a tropical climate. It involved the building of a rain-spray system simulator and worktops to support the experimental units. Three green roof modular-tray systems were studied, M-4L, M-17L and F-17L, with a mix of succulent plant species. The volume and delay of total runoff were sampled for both dry and wet soil-moisture stages for a simulated rainstorm of 155 mm/h and duration of 7 min, which approaches the value of the design rainfall of 10-year return-period and an observed rainfall event that caused flooding in 2011 in the area where the experiments took place. These results were compared to three bare-soil similar experimental units and with a control unit consisting of a corrugated fibre-cement roof sheet. In addition, classic rainfall-runoff parameterization, also important for catchment analysis, such as the rational coefficient method (C) and Curve Number (CN) were also applied. The results showed to be consistent. Overall, module F-17L presented the best performance with regard to both water retention and delay of total runoff. The average percentage of retention, considering all module types, was 58% of the total induced water volume, and the average delay for total runoff was approximately 12 min. In some tests, for the wet soil stage, the bare soil units performed better than the vegetated units. Soil moisture analysis revealed that considering the same period of time, the bare soil units lost more water through evaporation than the vegetated units through evapotranspiration. The average values of C for vegetated units for the wet and dry soil stages were equal to 0.66 and 0.27, respectively. The CN values were high, ranging from 88 to 97. Although fairly new on the market in Rio de Janeiro, modular-tray green roof systems are easy to build and hydrologically effective, mainly in the dry soil stages, performing as well as other extensive green roof systems. The values obtained for C and CN may be useful for the modelling of catchment rainfall and runoff especially for scenarios involving green roofs, as well as for establishing regulatory benchmarks towards more resilient and sustainable cities.

中文翻译：

---

模块化托盘绿色屋顶系统的水文性能提高特大城市对气候变化的适应能力

摘要 本文研究了模块化托盘绿色屋顶系统在大雨条件下（典型的热带气候国家）在雨水保持和总径流延迟方面的性能。它涉及建造雨喷系统模拟器和工作台以支持实验单元。研究了三种绿色屋顶模块化托盘系统，M-4L、M-17L 和 F-17L，混合了多汁植物物种。在模拟暴雨155 mm/h、持续时间7 min的干湿土壤水分阶段采样总径流体积和延迟，接近10年重现期设计降雨值和观察到的降雨事件导致 2011 年在进行实验的地区发生洪水。这些结果与三个类似的裸土实验单元和一个由波纹纤维水泥屋顶板组成的控制单元进行了比较。此外，还应用了经典的降雨径流参数化，对于集水区分析也很重要，例如合理系数法 (C) 和曲线数 (CN)。结果显示是一致的。总体而言，F-17L 模块在保水和总径流延迟方面表现最佳。考虑到所有模块类型，平均滞留百分比为总诱导水量的 58%，总径流的平均延迟约为 12 分钟。在一些测试中，对于湿土阶段，裸土单元比植被单元表现更好。土壤水分分析表明，考虑到同一时期，裸土单元通过蒸发损失的水分比植被单元通过蒸散损失更多。湿土和干土阶段植被单元的 C 平均值分别等于 0.66 和 0.27。CN 值很高，范围从 88 到 97。虽然在里约热内卢市场上相当新，但模块化托盘绿色屋顶系统易于建造且水文有效，主要在干土阶段，表现与其他广泛的绿色屋顶系统。获得的 C 和 CN 值可能有助于对集水区降雨和径流进行建模，特别是对于涉及绿色屋顶的场景，以及为更具弹性和可持续性的城市建立监管基准。湿土和干土阶段植被单元的 C 平均值分别等于 0.66 和 0.27。CN 值很高，范围从 88 到 97。虽然在里约热内卢市场上相当新，但模块化托盘绿色屋顶系统易于建造且水文有效，主要在干土阶段，表现与其他广泛的绿色屋顶系统。获得的 C 和 CN 值可能有助于对集水区降雨和径流进行建模，特别是对于涉及绿色屋顶的场景，以及为更具弹性和可持续性的城市建立监管基准。湿土和干土阶段植被单元的 C 平均值分别等于 0.66 和 0.27。CN 值很高，范围从 88 到 97。虽然在里约热内卢市场上相当新，但模块化托盘绿色屋顶系统易于建造且水文有效，主要在干土阶段，表现与其他广泛的绿色屋顶系统。获得的 C 和 CN 值可能有助于对集水区降雨和径流进行建模，特别是对于涉及绿色屋顶的场景，以及为更具弹性和可持续性的城市建立监管基准。模块化托盘绿色屋顶系统易于建造且在水文方面有效，主要在干燥土壤阶段，其性能与其他广泛的绿色屋顶系统一样。获得的 C 和 CN 值可能有助于对集水区降雨和径流进行建模，特别是对于涉及绿色屋顶的场景，以及为更具弹性和可持续性的城市建立监管基准。模块化托盘绿色屋顶系统易于建造且在水文方面有效，主要在干燥土壤阶段，其性能与其他广泛的绿色屋顶系统一样。获得的 C 和 CN 值可能有助于对集水区降雨和径流进行建模，特别是对于涉及绿色屋顶的场景，以及为更具弹性和可持续性的城市建立监管基准。