H43F-1018:
Green infrastructure retrofits on residential parcels: Ecohydrologic modeling for stormwater design
Thursday, 18 December 2014
Brian Miles, University of North Carolina at Chapel Hill, Geography, Chapel Hill, NC, United States and Lawrence E Band, Univ North Carolina, Chapel Hill, NC, United States
Abstract:
To meet water quality goals stormwater utilities and not-for-profit watershed organizations in the U.S. are working with citizens to design and implement green infrastructure on residential land. Green infrastructure, as an alternative and complement to traditional (grey) stormwater infrastructure, has the potential to contribute to multiple ecosystem benefits including stormwater volume reduction, carbon sequestration, urban heat island mitigation, and to provide amenities to residents. However, in small (1-10-km2) medium-density urban watersheds with heterogeneous land cover it is unclear whether stormwater retrofits on residential parcels significantly contributes to reduce stormwater volume at the watershed scale. In this paper, we seek to improve understanding of how small-scale redistribution of water at the parcel scale as part of green infrastructure implementation affects urban water budgets and stormwater volume across spatial scales. As study sites we use two medium-density headwater watersheds in Baltimore, MD and Durham, NC. We develop ecohydrology modeling experiments to evaluate the effectiveness of redirecting residential rooftop runoff to un-altered pervious surfaces and to engineered rain gardens to reduce stormwater runoff. As baselines for these experiments, we performed field surveys of residential rooftop hydrologic connectivity to adjacent impervious surfaces, and found low rates of connectivity. Through simulations of pervasive adoption of downspout disconnection to un-altered pervious areas or to rain garden stormwater control measures (SCM) in these catchments, we find that most parcel-scale changes in stormwater fate are attenuated at larger spatial scales and that neither SCM alone is likely to provide significant changes in streamflow at the watershed scale.