Planning considerations - Living Roofs in Integrated Urban Water Systems - page 76

(mm or inches), and I a is “considered the boundary between the storm size that
produces runoff and the storm size that produces no runoff” (NRCS 2004b:
10-5). Physically, it consists of canopy interception, initial iniltration, surface
depression storage and potentially ET (uSDA 1986). Equation 3.1 is valid for
storms greater than a minimum threshold given by P > I a , otherwise runoff is
considered zero.
A CN describes a generalized condition intended for design purposes, rather
than reproducing the runoff from a speciic historical (actual) rainfall record
(Hawkins et al. 2009). The CN gives an indication of runoff potential for non-
frozen soil; a high CN indicates high runoff potential while a low CN indicates
low runoff potential. The value of CN ranges between 39 and 98 for urban land
uses, increasing in magnitude with increasing impervious cover and/or decreasing
capability for a soil to iniltrate water from the surface. Watershed studies have
shown that the CN varies with rainfall depth (Hawkins et al. 2009), but in prac-
tice, only a single value is usually applied, as designers typically rely on informa-
tion provided by the NRCS (2004a: 9-9). The CN is selected from a table
combining descriptors of the watershed's condition including land use, its treat-
ment or condition, and characterization of the soil.
In the context of the CN method, a living roof “replaces” a runoff source area
described as an impervious surface (an unvegetated, conventional roof) which
would normally be a CN = 98. Since a living roof is considered an at- source
control, the value of the living roof's CN has to date been based on an assumed
hydrologic similarity to a natural surface in some regulatory guidance documents.
For example, until a living roof design manual was introduced in 2013 (Fassman-
Beck and Simcock 2013), Auckland's general SCM design manual assigned
CN = 61 (ARC 2003) which is equivalent to “open space in good condition” for
hydrologic soil group B soils (uSDA 1986). The Michigan LID Manual suggests
CN = 65 for extensive living roofs if the design rainfall event is up to three times
the moisture storage capacity of the living roof growing media. A CN for larger
design storms is not speciied (SEMCOG 2008).
The National Engineering Handbook indicates the most reliable means of estab-
lishing the CN is through the analysis of storm rainfall and resultant runoff data
(NRCS 1997), while ASCE (Hawkins et al. 2009) tests several methodologies for
determining a CN for sets of rainfall-runoff data . 7 In recent years, researchers have
been generating an ever-increasing amount of living roof performance characteri-
zation. While the number of data points pales in comparison to works cited by
ASCE (Hawkins et al. 2009), an opportunity nonetheless presents itself to begin to
assess potential CN s for living roofs applications based on observed data.
Few published studies have estimated CN s for living roofs from empirical data.
Applying a regression procedure based on TR-55, Carter and Rasmussen (2006)
derived a CN of 86 for a living roof in Georgia. In Michigan, Getter et al. (2007)
calculated CN s by the same method to be 84, 87, 89 and 90 for living roof test
plots with 2 percent, 7 percent, 15 percent and 25 percent slopes, respectively.
Based on simulated rainfall events in the laboratory and prototype living roofs