Estimation of Green Roof Stormwater Efficiency using Sensor-informed Models

Tuesday, July 23, 2013: 8:30 AM
Desert Salon 13-14 (Desert Springs J.W Marriott Resort )
Olyssa Starry , University of Maryland, College Park, MD
Andrew Ristvey , Wye Research and Education Center, University of Maryland, Queen Anne, MD
Steven Cohan , Plant Science and Landscape Architecture, University of Maryland, College Park, MD
John D. Lea-Cox , University of Maryland, College Park, MD
Rates of plant evapotranspiration (ET) are difficult to predict for Sedum species on green roofs, due to the high variability in ET rates, morphological differences and photosynthetic performance between species.. It may be possible to modify the standard FAO Penman Monteith reference equation used to predict ET for various crops species, but these model predictions must be calibrated against measurements of actual transpiration by each plant species.  In this presentation, we describe how a wireless sensor network was used to measure ET from four replicate 1.3 m2 platforms planted with three different Sedum species (S. album, S. kamtschaticum, and S. sexangulare) during 2011 and 2012.  Stormwater inputs, substrate volumetric moisture content, and runoff were monitored continuously from each platform throughout the two years of the study, every 15 minutes.  Environmental variables, i.e. air temperature, relative humidity, wind speed, solar radiation, and photosynthetic flux density were also collected every 5-minutes by the on-site weather station.  Root biomass, field capacity, and plant coverage were quantified seasonally for each planted platform.  In 2012, samples were also taken at three times during the year to determine leaf area for each species. Between March and November 2011, 985 mm (38 inches) of rain fell on the platforms.  Runoff totals were 851, 791, and 817L for the platforms planted in S.album, kampschaticum and sexangulare, respectively.  For storms during which > 50% of the water was retained by the green roof platforms, ANCOVA analysis revealed a treatment effect (P < 0.10) while accounting for variability in the data associated with storm size.  Ongoing data analysis will explain these findings by exploring relationships between runoff, morphological characteristics, and evapotranspiration and compare these among species for both 2011 and 2012.  This information will be used to parameterize and evaluate the FAO Penman Monteith equation for green roofs and to, for example, establish a crop coefficient for the species in this study.