Content area
Abstract
The objective of this dissertation is to improve understanding of how green infrastructure, including small-scale redistribution of water and changes to soils and vegetation, affects water budget and stormwater peaks and low flows across spatial scales in urbanized watersheds through the use of ecohydrology modeling. Given the expense of retrofitting existing development and building new stormwater infrastructure, ecohydrology models are indispensible tools for understanding the effects that stormwater control measures are likely to have on the hydrology of medium-density urbanized watersheds. However, these models require diverse geospatial datasets, the acquisition and preparation of which are labor intensive, yielding scientific workflows that are difficult to reproduce. In this research we develop improved representations of small-scale engineered surface drainage in the Regional Hydro-Ecological Simulation System (RHESSys) ecohydrology model, along with simulations of various residential area surface drainage configurations, as well as software tools, EcohydroLib and RHESSysWorkflows, that allow water scientists to rapidly develop reproducible ecohydrology modeling workflows.
We describe the design and demonstrate the use of EcohydroLib to acquire data needed to create RHESSys data preparation workflows using RHESSysWorkflows for three small headwater watersheds, two urban and one forested. Using baseline RHESSys models built using RHESSysWorkflows for two medium-density urbanized watersheds in Baltimore, MD and Durham, NC, we developed four scenarios of residential rooftop hydrologic connectivity to existing pervious surfaces or impervious surfaces drained by storm sewers. We found that disconnecting all single-family residential rooftops from nearby impervious surfaces results in decreased daily peak flows and increased daily baseflow, with a slight reduction in yearly streamflow. This result was only statistically significant for the Durham, NC watershed.
We developed additional modeling experiments to compare differences in water budget and stormwater volume across spatial scales when redirecting residential rooftop runoff to un-altered pervious surfaces versus engineered bioinfiltration rain gardens. Results show that at the watershed scale rooftop runoff redirected to rain gardens, even pervasively applied across two distinct watersheds, has only small, non-statistically significant effects (increasing base flow levels and decreasing storm flow peaks) when compared to scenarios with observed rates of residential rooftop connectivity to impervious surfaces determined through field surveys.
