Preprint: Simulated seasonal loads of total nitrogen and total phosphorus by major source from watersheds draining to Washington waters of the Salish Sea, 2005 through 2020

The U.S. Geological Survey (USGS) and the Washington State Department of Ecology (Ecology) have developed watershed models of seasonal load estimates of total nitrogen (TN) and total phosphorus (TP) discharging into the Washington waters of the Salish Sea from 2005 through 2020. The modeling approach used was dynamic SPARROW (SPAtially Referenced Regressions On Watershed attributes), a statistical-physical watershed modeling technique, initially applied at large spatial scales to represent long-term average stream loads throughout a stream network, refined here to estimate seasonal TN and TP loads across watersheds to clarify upstream contributions from discernable point and nonpoint sources delivered to marine waters at surface water confluences along the shoreline and quantify when, where, and why they were high or low. Upstream contributing sources included permitted treated wastewater facilities, crop fertilizer, animal feeding operations, septic systems, urban land and stormwater, atmospheric deposition (TN only), nitrogen fixation by Red Alder Alnus rubra trees (TN only), and background geologic material (TP only). Instream load magnitudes and their source compositions varied widely across watersheds, and even within each watershed, yet the largest loads typically occurred in the large rivers during winter and fall when streamflow was highest. Likewise, instream loads were typically lowest in summer during low streamflow, yet the relative instream aquatic decay was highest. The seasonal storage lag component of those nonpoint sources was estimated to contribute a quarter of the seasonal instream load during winter and fall high streamflow and sometimes half of the instream load during summer low streamflow. A key aspect of Ecology’s current Puget Sound Nutrient Source Reduction Project is consideration of upstream watershed contributions of nutrients to their marine-water discharge points. Simulated seasonal loads carried by streams to 63 river mouth marine discharge points 9 ranged by several orders-of-magnitude for both TN and TP due to the spatial and seasonal differences in hydrologic flows, magnitude and timing of contributing sources, and in-stream decay. The Snohomish and Skagit Rivers discharged the largest TN and TP loads, yet the Samish River was shown to have some of the highest TN and TP yields and concentrations. Additionally, a reference scenario was developed to provide an estimate of the pre-industrial local and regional loads.