Evaluating Instream Restoration Effectiveness In Reducing Nitrogen Export From An Urban Catchment With A Data-Model Approach

Urbanization increases stormwater runoff into streams, resulting in channel erosion, and increases in sediment and nutrient delivery to receiving water bodies. Stream restoration is widely used as a Best Management Practice to stabilize banks and reduce sediment and nutrient loads. While most instream nutrient retention measurements are often limited to low flow conditions, most of the nutrient load is mobilized at high stream flows in urban settings. We, therefore, use a process-based stream ecosystem model in conjunction with measurements at low flows and focus on estimation of stream nitrogen retention over the full streamflow distribution. The model provides a theoretical framework to evaluate the geomorphic, hydrologic, and ecological factors that are manipulated by stream restoration, and drive nitrogen retention. We set a model for a pool-riffle sequence restored stream (190 m) in Baltimore County, Maryland and calibrated the model to the in situ measured primary production (Nash-Sutcliffe model efficiency coefficient [NSE] NSE = 0.89), respiration (NSE = 0.74), and nitrate uptake lengths (R-2 = 0.88). At the daily scale, simulations showed low nitrogen retention during high flows due to high transport rates, mobilization of stored hyporheic nitrogen, and scouring of periphyton biomass. This result underscores the need to reduce contributing watershed runoff flashiness to promote aquatic nutrient cycling and retention. At monthly and yearly time scale, model predicted a higher percent reduction in summer than in winter and estimated 5.7%-9.5% of annual nitrate reductions. While the model was tested in a pool-riffle sequence restoration design, the approach can be adapted to evaluate a range of channel restoration design characteristics, and the effects of upland watershed restoration to mitigate stormwater loading through both restored and unrestored streams.