Curve Number and Runoff Coefficients for Extensive Living Roofs

Living roofs are a green infrastructure (GI)/low-impact development (LID) stormwater control measure (SCM) increasingly drawing worldwide attention. Despite substantial performance evidence in the literature, the lack of a curve number (CN) or volumetric runoff coefficient ($C_v$) to apply to prescribed methodologies for planning and regulatory submissions may be perceived as a barrier for implementation. Paired rainfall–runoff data were analyzed for up to 21 living roofs with varying configurations and in different climates from studies identified in the literature and previously-unpublished data. Frequency analysis of empirical performance evidence from 14 living roofs indicates that meaningful runoff is not generated from the majority of small rainfall events. Where planning requires the use of the CN method, a step function is suggested: (1) runoff $\text{volume}=0$ for design rainfall events up to 20–30 mm, if appropriate moisture storage capacity is provided by the substrate; (2) runoff volume is determined with $\mathrm{CN}=84$ for larger rainfall events, or for events that exceed the actual moisture storage capacity. $C_v$ increases with rainfall depth ($P$), and may be reasonably predicted for 16 living roofs by $C_v=a*\exp (b/P)$, where regression coefficients $a$ and $b$ were empirically determined for each climate zone. CN or $C_v$ values are considered best fit, but they are generally poor representations of actual living roof hydrology. Further work is required to develop living roof specific, verified continuous simulation computer-modeling techniques and to quantify the role of the plants in stormwater control.