About this indicator
Excess nitrogen and phosphorus loading impacts not only local waters, but also downstream waterbodies and coastal systems including the Chesapeake Bay, the Great Lakes, the Gulf of Mexico, and Puget Sound. The data in this indicator table are based on output from the United States Geological Survey’s (USGS) Spatially Referenced Regression On Watershed Attributes (SPARROW) models. SPARROW is a watershed modeling tool which, based on a mass-balance approach, estimates the excess amounts (i.e., amounts beyond assimilative capacity) of nitrogen and phosphorus exported from watersheds and delivered to downstream waterbodies. The models relate in-stream water-quality measurements of nutrients taken at a network of monitoring stations to spatially referenced attributes of the corresponding watersheds, such as nutrient sources and environmental factors that affect rates of delivery to streams, as well as in-stream processing of nutrients.

Estimated state-wide total nitrogen (TN) and total phosphorus (TP) loads and yields, as predicted by the 2002 regional USGS SPARROW models.

Note - These data reflect the excess amounts ('loads') of TN and TP generated beyond the assimilative capacity of each states' watersheds. Net incremental TN or TP load is the sum of the area-weighted⁺⁺ incremental TN or TP loads from all catchments (drainage basins) within a state that are covered by one or more regional SPARROW models. Also shown is the average yields for each state, which is the net incremental TN or TP load within a state divided by the area of the state with SPARROW model coverage.

* Denotes states for which 2002 SPARROW model data were unavailable for most or all of the state area.

** % of state area covered by respective 2002 SPARROW model(s): Colorado - 56%; Florida - 75%; Oregon - 93%; Texas - 82%; Wyoming - 78%.  Values shown only reflect area covered. All other states are covered within the range of 97-100% of state area.

⁺⁺Where only a portion of a catchment was within state boundaries, the contribution from that catchment was calculated as the incremental load from the catchment scaled by the portion of the catchment within the state's boundaries.

Source: USGS' SPARROW Decision Support Tool.

Download the Loads data table (excel)(2 pp, 26 K)

Top of Page

Sources of data
1. U.S. Geological Survey. Spatially Referenced Regression On Watershed attributes (SPARROW) Decision Support System.

2. NHDPlus Version 01 catchment polygons for Production Unit 1 (New England) and 2 (Mid-Atlantic).

3. U.S. Geological Survey. A Digital Hydrologic Network Supporting NAWQA MRB SPARROW Modeling--MRB_E2RF1WS (MRB_E2RF1 catchments).

4. U.S. Census Bureau, Geography Division. State boundary polygon dataset from the 2010 Census work. Created September 12, 2011, revised January 31, 2012.

Data source information
State-level estimates of TN and TP loads and yields were made as follows:

• USGS 2002 regional SPARROW model outputs for TN and TP incremental yields (kg/yr/km²) were obtained as CSV files from the USGS SPARROW Decision Support System (DSS) for each of the six geographical regions (referred to as ‘Major River Basins’) for which USGS has generated 2002 estimates. Data files were merged in Microsoft ACCESS and converted to DBF (DataBase Files) format for use with GIS (Geographical Information System) software.
• Geospatial files (i.e., GIS polygons or shapefiles) representing the hydrologic networks used in the 2002 SPARROW models were obtained from USGS. The New England and Mid-Atlantic (Major River Basin 1 or MRB1) model was based on NHD (National Hydrography Dataset) Plus catchments, whereas all other 2002 models were based on MRB_E2RF1 (Major River Basin_Enhanced River Reach File) catchments.
• The hydrologic network data and associated DSS output from all of the models were joined in ESRI’s ArcGIS v.10 to generate two seamless grids of incremental yield estimates (kg/yr/km²): one for TN and one for TP. The DSS output for each catchment was linked to the hydrologic network data using the “ID” attribute field in the DSS output and the appropriate unique identifier for the hydrologic network file (“GRID_CODE” attribute field for NHDPlus catchments, and “Value” attribute field for MRB_E2RF1).
  • Note: Because NHDPlus and MRB_E2RF1 catchments are of significantly different scales, the model extents/boundaries between the model using NHDPlus catchments (New England and Mid-Atlantic model or “MRB1”) and adjacent models using MRB_E2RF1 catchments (Southeastern US or “MRB2” and Upper Midwest or “MRB3”) have small overlaps and gaps. Overlapping areas were assigned the yield estimates from the MRB2 or MRB3 model by using the Erase tool of Xtools Pro (an add-on toolbar for ArcGIS) to remove the yield estimates from the MRB1 model. No estimates of incremental yields were assigned to the gaps. (However, note that these gaps are accounted for in the percent of state area covered by the respective model(s), which was used to determine incremental yields.)
• Xtools Pro was used to calculate the area of all model catchments and the ArcGIS Intersect tool was used—with the state boundaries file from the 2010 U.S. Census—to limit the corresponding TN or TP incremental yields to model output data from catchment areas within state boundaries. This was done so that incremental loads for catchments that cover more than one state were apportioned using an area-weighted approach. All data were then exported into Microsoft Excel for post-GIS processing.
• For each catchment within state boundaries, incremental load (kg/yr) was determined by multiplying the incremental yield (kg/yr/km²) by the area of the catchment (km²) within state boundaries (determined in previous step). Net incremental load is the sum of all incremental loads within state boundaries.
• Average yield for all catchment areas within the boundaries of each state was determined by dividing the net incremental load of TN or TP within state boundaries by the total state area covered by one or more of the regional models.

Note - Any reference to trade names or commercial products is for informational purposes only and does not constitute any endorsement or recommendation for use. US EPA and its employees do not endorse any commercial products, services or enterprises.

What to consider when using these data
With the exception of Florida and some western states, 2002 SPARROW model coverage of states is high, in the range of 97% - 100% of state area. In that coverage range, states without 100% coverage are generally coastal states and areas not covered by DSS output are generally coastal regions. Although 75% of Florida is covered in the 2002 South Atlantic Gulf and Tennessee (region 2) SPARROW model, significant portions of central Florida and all of south Florida are excluded. For Texas, Colorado and Wyoming, a significant and likely representative portion of the state (82%, 56%, and 78%, respectively) is covered in at least one of the 2002 regional SPARROW models. Because of the lack of complete or near complete coverage of all states, loads and yields of TN and TP currently covered in the 2002 models. For states with a very low or zero coverage by the 2002 regional models, estimates of TN and TP loads and yields were not made (indicated by ‘No data’ or ‘Insufficient data’ in the data table). SPARROW Modeling: estimating nutrient, sediment and dissolved solids transport.

Like all models, the SPARROW models have uncertainty associated with them. To minimize uncertainty (or the unexplained differences between measured and predicted values, as measured by each model’s coefficient of determination [R² value] and root mean squared error), USGS developed their 2002 regional SPARROW models using water quality monitoring sites for model calibration, and geospatial datasets on source contributions and delivery factors such as basin characteristics and water flow. In all cases, the load R² values for both TN and TP models are ≥ 0.84, and 8 of the 12 models have load R² values that are ≥ 0.90 (Preston et al. 2011). Additionally, on the DSS website, users can view, on a regional model basis, the standard error of the estimated incremental TN and TP loads for each catchment or reach. Due to complications associated with translating uncertainty at the level of stream reaches to an estimate of uncertainty for the overall load and yield per state, state-wide values of uncertainty have not been calculated.

References and links to other data sources
1. Preston, S.D., Alexander, R.B., Schwarz, G.E. and Crawford, C.G. 2011. Factors affecting stream nutrient loads: a synthesis of regional SPARROW model results for the continental United States. JAWRA Journal of the American Water Resources Association. Vol. 47, pp. 891–915

2. Preston, S.D., Alexander, R.B., and Woodside, M.D. 2011. Regional assessments of the Nation's water quality - Improved understanding of stream nutrient sources through enhanced modeling capabilities: U.S. Geological Survey Fact Sheet 2011-3114.

3. Committee on Environment and Natural Resources. 2010. Scientific Assessment of Hypoxia in U.S. Coastal Waters. Interagency Working Group on Harmful Algal Blooms, Hypoxia, and Human Health of the Joint Subcommittee on Ocean Science and Technology. Washington, DC.

4. Diaz, R. J., R. Rosenberg. 2008. Spreading dead zones and consequences for marine ecosystems. Science. Vol. 321, pp. 926-929.

5. Bricker, S., B. Longstaff, W. Dennison, A. Jones, K. Boicourt, C. Wicks, and J. Woerner. 2007. Effects of Nutrient Enrichment in the Nation's Estuaries: A Decade of Change. NOAA Coastal Ocean Program Decision Analysis Series No. 26. National Centers for Coastal Ocean Science, Silver Spring, MD.

6. Puckett, L.J. 1994. Nonpoint and point sources of nitrogen in major watersheds of the United States: USGS Water-Resources Investigations Report 94-4001.

7. U.S. Geological Survey. 2011. SPARROW Modeling: estimating nutrient, sediment and dissolved solids transport.

8. National Oceanic and Atmospheric Administration. National Centers for Coastal Ocean Science: Learning more about estuarine eutrophication.

Top of Page