![[logo] US EPA](../gif/logo_epaseal.gif){kind=logo}
New England Sparrow Model - Potentially Useful Infomration for Designing a Stream Water-Quality Network
Richard B. Moore and Keith W. Robinson
U.S. Geological Survey, Pembroke, New Hampshire
The New England SPAtially Referenced Regressions On Watershed Attributes models (New England SPARROW) are spatially detailed regression models that relate annual nutrient stream loads (phosphorus and nitrogen) to potential sources and watershed characteristics. These statistical relations are used to predict nutrient loads in unmonitored streams. The U.S. Geological Survey (USGS), in cooperation with the U.S. Environmental Protection Agency (USEPA) and the New England Interstate Water Pollution Control Commission (NEWIPCC), has developed these water-quality models to assist in the determination of regional total maximum daily loads (TMDL) and in the development of regional nutrient criteria options for New England.
The New England SPARROW models are built using a hydrologic network of 42,000 stream reaches in the 1:100,000-scale National Hydrography Dataset (NHD) and their associated watershed characteristics. Statistically significant predictor variables used in the models include nutrient source data from wastewater discharges, various land use types, and atmospheric deposition (for nitrogen only).
The model results, which are linked to the NHD reaches, provide estimates and confidence intervals of loads, area-weighted yields, flow-weighted mean annual concentrations, and sources and downstream movement of nutrients. This information can be useful for determining appropriate monitoring strategies and has been used to select monitoring sites in an upper Connecticut River Basin study. Other approaches include using the prediction intervals to identify stream reaches where additional monitoring would be beneficial. Sampling schemes could be designed to target stream reaches of greatest uncertainty relative to critical concentrations of nutrients.
Keywords: SPARROW, nutrients, regression model, NHD, water-quality network