Modeling Nitrogen Processing in Northeast US River Networks

Due to increased nitrogen (N) pollution from anthropogenic sources, the need for aquatic ecosystem services such as N removal has also increased. River networks provide a buffering mechanism that retains or removes anthropogenic N inputs. However, the effectiveness of N removal in rivers may decline with increased loading and, consequently, excess N is eventually delivered to estuaries. We used a spatially distributed river network N removal model developed within the Framework for Aquatic Modeling in the Earth System (FrAMES) to examine the geography of N removal capacity of Northeast river systems under various land use and climate conditions. FrAMES accounts for accumulation and routing of runoff, water temperatures, and serial biogeochemical processing using reactivity derived from the Lotic Intersite Nitrogen Experiment (LINX2). Nonpoint N loading is driven by empirical relationships with land cover developed from previous research in Northeast watersheds. Point source N loading from wastewater treatment plants is estimated as a function of the population served and the volume of water discharged. We tested model results using historical USGS discharge data and N data from historical grab samples and recently initiated continuous measurements from in-situ aquatic sensors. Model results for major Northeast watersheds illustrate hot spots of ecosystem service activity (i.e. N removal) using high-resolution maps and basin profiles. As expected, N loading increases with increasing suburban or agricultural land use area. Network scale N removal is highest during summer and autumn when discharge is low and river temperatures are high. N removal as the % of N loading increases with catchment size and decreases with increasing N loading, suburban land use, or agricultural land use. Catchments experiencing the highest network scale N removal generally have N inputs (both point and non-point sources) located in lower order streams. Model results can be used to better predict nutrient loading to the coastal ocean across a broad range of current and future climate variability.