The role of living roofs in holistic stormwater management systems - Living Roofs in Integrated Urban Water Systems - page 36

exceeds evaporation from an unplanted growing medium. During very dry
periods, ET occurs at a rate similar to evaporation from bare surfaces.
Studies that investigated ET on shorter timescales (i.e., hourly measurements)
show that ET can be affected by the environmental conditions as they change
throughout the day (Rezaei 2005; Starry 2013; Voyde et al. 2010; Yio et al.
2013). This means that if the sun comes out even for a few hours between rain-
fall bursts, and other ambient conditions are not oppressive (relative humidity can
be a signiicant inhibitor of ET), some small amount of storage capacity may be
recovered on a relatively short timescale. However, during a storm event, the
actual magnitude of hourly ET is generally quite a bit smaller than the typical
amount of rainfall (Stovin et al. 2013).
Most studies agree that plants such as succulents adjust their metabolism as
water becomes limiting and/or plants become stressed to conserve water for future
use. This is one of the primary adaptations of many succulent species that render
them generally well-adapted to the droughty living roof environment. In some cases,
this behavior implies Crassulacean Acid Metabolism (CAM), in which plants take up
CO 2 at night, to metabolize it during the day with stomates closed thereby reducing
plant water loss (Wolf 1960). The opposite is known as C3 photosynthesis, whereby
plants take up CO 2 during daylight with open stomates, thereby rendering the plant
more susceptible to transpiration water losses (due to higher energy [radiation] from
sunlight). Some individual sedum species exhibit CAM exclusively; some do not
exhibit it at all, while other species have the ability to shift between CAM and C3
(Berghage et al. 2007; Starry et al. 2014; Voyde 2011). Starry et al. (2014) explore in
detail factors contributing to CAM in sedum species, and a review of the plant
metabolism literature in the living roof context. From a stormwater engineering
standpoint, quantifying the rate at which ET decreases as water becomes limiting is
more important than the observance of a speciic metabolic pathway.
Daily measurements of living roof ET vary amongst studies of sedum-planted
extensive living roof assemblies. Voyde (2011) measured maximum ET values for
days when water was readily available at 1.5-5.4 mm/day whereas days with
water-limited conditions resulted in 0.0-0.7 mm/day under both laboratory and
ield conditions in Auckland. Over the course of one year's monitoring of a ield
site in New York, DiGiovanni et al. (2013) reported average daily ET ranging from
approximately 0.5 mm/day during winter months to a maximum of about
3.5 mm/day during June. In Shefield, UK, Berretta et al. (2014) identiied average
moisture loss rates of approximately 1.5-1.8 mm/day for dry periods in May and
July, when the growing media's initial moisture content was high (i.e., good
water availability), versus average moisture loss rates of 0.1-0.8 mm/day when
the initial moisture content was low in March, April and July. It is important to
recognize that the average ET rate reported for a dry period depends on the
duration of the dry period; long interevent times allow for more drying, but the
actual ET rate will be low toward the end of that period contributing to an overall
lower average metric. Conversely, a short interevent time yields a higher average
rate as relatively more water is available for the period (Berretta et al. 2014).