Green roofs can be an effective tool for sustainable urban drainage, since they reduce and retain runoff by delaying its peak. Most studies analysing the retention capacity of green roofs are usually referred to a specific place and roof condition and do not consider the possibility that the roof could be partially pre-filled from previous rainfalls at the beginning of the given event. The aim of this paper is to develop an analytical probabilistic approach to evaluate green roof performance for stormwater control in terms of runoff that could be applied for different sites and climate conditions. To this end, the possibility that the green roof retention capacity could not be completely available owing to pre-filling from previous rainfall events has been considered and equations for an optimum green roof design, relating the runoff average return interval to the water retention capacity, have been proposed. The influence of parameters affecting the runoff process has been examined in depth and a case study to test the goodness of fit of the resulting equations has been developed.

绿化屋顶可以通过延迟峰值来减少并保持径流，因此可以成为可持续城市排水的有效工具。大多数分析绿色屋顶保持能力的研究通常都针对特定的位置和屋顶条件，并且没有考虑在给定事件开始时屋顶可能会因先前的降雨而被部分预填充的可能性。本文的目的是开发一种分析概率方法，以评估径流方面可用于不同地点和气候条件的雨水控制的绿化屋顶性能。为此，已经考虑了由于先前降雨事件的预填充而无法完全利用屋顶绿化的能力，并且考虑了最佳屋顶绿化设计的方程式，已经提出将径流平均返回间隔与保水能力联系起来。深入研究了影响径流过程的参数的影响，并开发了一个案例研究来检验所得方程的拟合优度。