Watershed-Scale Effects of Floodplain Restoration on Nitrate Removal
Morgan A. Oehler
Virginia Tech
Blacksburg, VA
Authors:
-- Morgan A. Oehler, Civil and Environmental Engineering, Virginia Tech, Blacksburg VA
-- Durelle T. Scott, Biological Systems Engineering, Virginia Tech, Blacksburg VA
-- Erich T. Hester, Civil and Environmental Engineering, Virginia Tech, Blacksburg VA
& AAAS Fellow with Water Power Technologies Office, U.S. Dept. of Energy, Washington DC
Excess nutrients from human activities such as agriculture and urbanization threaten downstream water quality and ecosystems. Enhancing river water exchange with floodplain vegetation and soils through floodplain restoration can reduce nutrients reaching downstream waterbodies, yet this potential has been poorly quantified at the watershed scale. We simulated the effect of cumulative and spatially varying floodplain restoration on dissolved nitrate removal using the U. S. Army Corps of Engineers Hydrologic Engineering Center’s River Analysis System (HEC-RAS) and R. First, we modeled a synthetic 4th-order watershed in HEC-RAS using average stream geometry and hydrology for the Virginia Piedmont with storms ranging in size from the 2-year down to monthly discharges. We modeled nitrate removal in the floodplains using a mass balance, zero-order removal approach in R. Nitrate removal was greatest at the location of restoration with relatively little effect downstream. Nitrate removal was limited by inundation time in lower order streams and inundation area in higher order streams. Although percent watershed nitrate removal increased with restoration length and river order, the significance of removal was strongly dependent on the assumed nitrate removal rate. For example, when we varied the percent of the 4th-order watershed restored, the median removal rate yielded insignificant watershed nitrate load reduction (maximum of 0.9%), whereas the 90th-percentile rate yielded reductions up to 19%. When we scaled our results to the annual timescale and compared them to results for hyporheic restoration from a prior study, we found hyporheic restoration removed substantially more watershed nitrate than floodplain restoration. We then applied our floodplain modeling approach to the Gwynn’s Falls watershed in Baltimore, MD, using georeferenced HEC-RAS models from the Maryland Department of the Environment (MDE) Flood Risk Application. We found that the challenges of applying our unsteady analysis to a real watershed included availability and quality of site-specific data, and calibrating flows for complex geometries. Overall, our results highlight the potential value and limitations of floodplain restoration in reducing nitrate exports at the watershed scale and provide practical insight for use of our simulation techniques in applied watershed water quality management.
About Morgan A. Oehler
Morgan Oehler grew up around the Chesapeake Bay. Inspired by the natural systems around her she became passionate about river systems and watershed processes. Morgan has a master's degree in civil and environmental engineering from Virginia Tech and a bachelor’s degree from the University of Maine. Her thesis research involved modeling the cumulative and spatially varying effects of floodplain restoration on nitrate removal at the watershed scale. Morgan is pursuing her career as a water resources engineer in stream restoration and in her free time enjoys open water swimming.