Assessing the Medium-Term Changes in Hydrological Performance of Green Roofs: The Influence of Precipitation Severity

As urbanization and climate change continue to pose significant challenges for cities worldwide, green roofs (GRs) has emerged as a viable sustainable solution for supporting traditional infrastructure in managing stormwater runoff. Although their hydrological behavior has been sufficiently documented in literature, conflicting results emerge regarding the potential variations in their retention capacity (RC) over the medium and long-term. Based on preliminary investigations, this research aimed at assessing medium-term changes in the hydrological performance of two experimental GRs (GR1 and GR2), further investigating the potential role played by precipitation severity. The GRs, located in Southern Italy and consisting of three layers (vegetation, substrate and drainage), were set up in 2017 and monitored for two operational periods, 2017-2019 and 2022-2023. The measurements gathered between 2017 and 2019 provide valuable insights into the initial performance of the GRs and their ability to retain water during the early years of operation. Data collected in 2022 and 2023 instead reflect the retention capacity of the GRs after a few years of operation. A total of 29 mild precipitation events were collected during both periods and for both GRs, detecting from the monitoring data their cumulative precipitation (P) and runoff (R) with the objective of assessing the RC (RC = 1 - R/P). Based on the preliminary findings, it appears that there is an overall decline in the RC for both GR1 and GR2, without significant differences between the two. The Aging Indexes (AI) were calculated for GR1 and GR2, representing the average reduction of the runoff coefficient (RC) over time. GR1, which has a drainage layer composed of expanded clay, exhibited an AI of 12%. On the other hand, GR2, characterized by a drainage layer made of MODI' plastic panel filled with expanded clay, exhibited a slightly higher AI of 13%. Further analysis revealed that within each dataset, two groups were identified based on a threshold determined by the growth coefficient g(T) of the precipitation events. For the group of events with g(T) values above 0.12 (sample size of 14), the AI values were 15% and 16% for GR1 and GR2, respectively. On the other hand, the group of events with g(T) values equal to or lower than 0.12 (sample size of 15) experienced AI values of 10% and 11% for GR1 and GR2, respectively. These findings suggest that as the growth coefficient g(T) increases, indicating higher return periods T, the AI and consequently the reduction in hydrological performance of GRs also increase. The highly possible increase in the future of extreme precipitations would pose a considerable limit to the spread of this kind of sustainable drainage infrastructures. However, additional modeling investigations focused at detecting the effects of alternative GRs designs and materials on their long-lasting average hydrological performance would be essential for making informed decisions and investments.