Hydrology of an extensive living roof under sub-tropical climate conditions in Auckland, New Zealand Emily Voyde a,⇑ , Elizabeth Fassman a , Robyn Simcock b a Department of Civil and Environmental Engineering, University of Auckland, New Zealand b Landcare Research, Auckland, New Zealand article info Article history: Received 9 April 2010 Received in revised form 30 August 2010 Accepted 17 September 2010 This manuscript was handled by Konstantine P. Georgakakos, Editor-in-Chief, with the assistance of Efrat Morin, Associate Editor Keywords: Living roof Green roof Stormwater management Environmental engineering New Zealand summary This paper presents field monitoring results from a 235 m 2 , extensive living roof in Auckland, New Zealand (NZ). The extent of stormwater control is quantified by comparing three different substrate types (Pumice, Zeolite and Expanded Clay, all pumice based but named for their distinguishing compo- nents) at two different substrate depths (50 and 70 mm) in a side-by-side comparison. No statistically significant differences in runoff response were found between the three substrate types tested or the two different depths. The cumulative retention efficiency of the living roof was 66% based on 12 months of continuous monitoring. On an event basis, the living roof demonstrated reductions in both volume and peak flow rates regardless of the rainfall and climatic characteristics. The living roof retained a median of 82% of rainfall received per rainfall event, with a median peak flow reduction of 93% compared to rainfall intensity. The hydrologic response of a living roof is controlled by multiple parameters such as rain depth, rain intensity, climatic variables and antecedent dry days. Detailed analysis indicates that antecedent dry days have the greatest influence on retention. Seasonal differences do not influence runoff response; living roofs will effectively moderate runoff hydrology year round in Auckland’s sub-tropical climate. Ó 2010 Elsevier B.V. All rights reserved. 1. Introduction Stormwater runoff from urban rooftops makes a significant con- tribution to urban water quality problems and nuisance flooding. Any technique that reduces the volume and rate of roof runoff has the potential to contribute to improved stormwater manage- ment. A living roof (also commonly referred to as green roof, veg- etated roof, eco-roof, roof garden or landscape over structure (Weiler and Scholz-Barth, 2009)) provides an opportunity to mitigate stormwater runoff at the source. It typically consists of multiple layers (Fig. 1), each of which plays an important role in the overall system function. Two main living roof categories are defined based on substrate depth: intensive (deeper) and extensive (shallower). The depth defining the transition from extensive to intensive living roof varies from 100 mm to 200 mm depending on the source (FLL, 2002; Getter and Rowe, 2006, 2008; Oberndorfer et al., 2007). The thin depth of substrate for an extensive living roof limits the plant height and species that can be grown without irrigation (Dunnett and Kingsbury, 2004; Snodgrass and Snodgrass, 2006), but extensive living roofs are better suited for retrofit application due to lighter weight requirements compared to their intensive counterparts. Plants play a significant role in the func- tionality of a living roof for stormwater management. Dependant on plant type, season and water availability, plants can contribute 20–48% of total evapotranspiration via the process of transpiration, thus aiding storage recovery within the substrate (Berghage et al., 2007; Rezaei and Jarrett, 2006; VanWoert et al., 2005a; Voyde et al., 2010). Additional features may be included within the living roof design to enhance stormwater function and vegetation health: drainage boards or mats with cups designed to store water, mois- ture retention mats (geotextiles), substrate layers with contrasting textures, surface mulches and increased distance to downpipe outlets. Field monitoring of full scale (field installation) extensive living roofs show a substantial contribution to site runoff volume control with 50–78% of precipitation retained over extended periods of data collection (Berghage et al., 2009; Carter and Rasmussen, 2006; Hutchinson et al., 2003; Mentens et al., 2006; Moran et al., 2005; Villarreal and Bengtsson, 2005). Shorter duration studies (<6 months) tend to report lower precipitation retention (DeNardo et al., 2005; Liu, 2003; Moran et al., 2005; Villarreal, 2007). This is likely a function of the algorithm used to calculate %-reduction; fewer rainfall events are captured (less depth) and a single large event has greater potential to skew the overall reduction. Villarreal (2007) found that lower retention volumes were observed for 0022-1694/$ - see front matter Ó 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.jhydrol.2010.09.013 ⇑ Corresponding author. Address: Department of Civil and Environmental Engineering, University of Auckland, Private Bag 92019, Auckland Mail Centre, Auckland 1142, New Zealand. Tel.: +64 9 3737599x89555; fax: +64 9 373 7462. E-mail address: evoy001@aucklanduni.ac.nz (E. Voyde). Journal of Hydrology 394 (2010) 384–395 Contents lists available at ScienceDirect Journal of Hydrology journal homepage: www.elsevier.com/locate/jhydrol