Role of Precipitation Variability in Interannual Variation of Nitrogen Loading in Streams in the Continental U.S.

Abstract:

Although nitrogen application in the form of fertilizers increases food production, excess nitrogen is harmful to the environment and to human well-being. Most watersheds have limited or no observational water quality data for estimating nitrogen load. Models predicting nitrogen load for various watersheds are utilized under such circumstances. However, existing model-based estimates of nitrogen loading to streams are limited either spatially (single river basin or small regional scale) or temporally (a single year). For developing strategies for reducing nutrient loads it is crucial to have knowledge of nitrogen loading for various regions of the country and over several time periods to understand the impact of precipitation variability on nitrogen loading. In this study we propose a parsimonious empirical model based on precipitation, land use and nitrogen input to the watershed and find that it explains 68% (76%) of the observed variability in annual total nitrogen (TN) flux (log(TN flux)) across 242 watersheds-years used in the model calibration. We use this model to create estimates of TN flux covering all HUC8 watersheds within the Continental United States and spanning the years 1987 to 2007. Results show that although net anthropogenic nitrogen input is the primary explanatory factor for average nutrient loading across years, the interannual variability in TN flux is primarily attributable to variability in precipitation for 86% of watersheds, which together represent 98% of the estimated TN flux for the continental U.S. These results crystallize the challenges associated with managing water quality impacts for systems where nitrogen input is the primary contributor to TN flux but precipitation controls its year-to-year variability.