Spatially Referenced Models of Streamflow and Nitrogen, Phosphorus, and Suspended-Sediment Loads in Streams of the Midwestern United States

Scientific Investigations Report 2019-5114
National Water Quality Program
By:  and 

Links

  • Document: Report (43.7 MB pdf)
  • Application Site: Mapping application – Online mapping tool to explore 2012 SPARROW Models
  • Companion Files:
    • SIR 2019–5106 – Spatially Referenced Models of Streamflow and Nitrogen, Phosphorus, and Suspended-Sediment Loads in Streams of the Southwestern United States
    • SIR 2019–5112 – Spatially Referenced Models of Streamflow and Nitrogen, Phosphorus, and Suspended-Sediment Loads in Streams of the Pacific Region of the United States
    • SIR 2019–5118 – Spatially Referenced Models of Streamflow and Nitrogen, Phosphorus, and Suspended-Sediment Loads in Streams of the Northeastern United States
    • SIR 2019–5135 – Spatially Referenced Models of Streamflow and Nitrogen, Phosphorus, and Suspended-Sediment Loads in Streams of the Southeastern United States
  • Data Release: USGS data release – SPARROW model inputs and simulated streamflow, nutrient and suspended-sediment loads in streams of the Midwestern United States, 2012 base year
  • Download citation as: RIS | Dublin Core

Abstract

In this report, SPAtially Referenced Regression On Watershed attributes (SPARROW) models developed to describe long-term (2000–14) mean-annual streamflow, total nitrogen (TN), total phosphorus (TP), and suspended-sediment (SS) transport in streams of the Midwestern part of the United States (the Mississippi River, Great Lakes, and Red River of the North Basins) are described. The nutrient and suspended-sediment models have a base year of 2012, which means they were developed based on source inputs and management practices similar to those existing during or near 2012 and average hydrological conditions detrended to 2012 (2000–14), whereas the streamflow model has base years of 2000–14, which means it was developed based on the average input precipitation minus actual evapotranspiration from 2000 to 2014. In developing the models, several updates and improvements were made to the data inputs and statistical approaches used to calibrate/develop the models from those used in the previous 2002 SPARROW models. The 2012 SPARROW models were constructed using a higher resolution stream network, which resulted in a mean catchment size of 2.7 square kilometers compared to 480 square kilometers in the 2002 models; more detailed and updated wastewater treatment plant contribution estimates; inputs from background phosphorus sources that were not included in the 2002 model; and more accurate loads for calibration that were computed using a modified Beale ratio-estimator technique whenever no trend in load was determined. Statistical approaches were added to compensate for the unequal effect of each monitoring site during the calibration process by adjusting for the fraction of the basin included in other upstream monitored sites (nested share) and thinning the calibration sites if a negative statistical correlation between nearby sites was determined.

Results from 2012 SPARROW models describe how much of each water, TN, TP, and SS source was delivered to the stream network, and the major landscape factors that affected their delivery. Atmospheric deposition and natural (background) sources of TN and TP, respectively, were the dominant sources in anthropogenically unaffected areas (especially in the Rocky Mountains and north-central areas of the Midwest), whereas fertilizers, manure, and fixation were dominant sources in agricultural areas, especially in the Corn Belt and near the Mississippi River. Urban sources of TN and TP were typically localized, but they were still important for some large areas, especially the Lake Erie Basin. All of the land-to-water delivery variables in the nutrient and sediment SPARROW models, such as runoff, soil erodibility, basin slope, and the amount of tile drains, are commonly included in process-driven models. In the SPARROW TN and TP models, best management practices (BMPs) reduced the delivery of these nutrients to streams.

Long-term mean-annual flows and nutrient and sediment loads were simulated in streams throughout the Midwest. The simulated flows from the SPARROW flow model were used in the SPARROW TN, TP, and SS models to help describe nutrient and sediment transport from the watershed and through the stream network. Outputs from the TN, TP, and SS models describe loads and yields of these constituents throughout the Midwest, and from major drainage basins throughout the Midwest. Highest TN, TP, and SS yields and delivered yields were from the Lake Erie, Ohio River, Upper Mississippi River, and Lower Mississippi River Basins, whereas lowest yields were spread over most other areas. Losses during downstream delivery resulted in part of the TN, TP, and SS that reach the stream network not reaching the downstream receiving bodies: 14, 15, and 28 percent of the TN, TP, and SS, respectively, are lost during delivery to the Great Lakes and 19, 23, and 52 percent of the TN, TP, and SS, respectively, are lost during delivery to the Gulf of Mexico. The largest losses of nutrients and sediments during transport were in the Missouri and Arkansas River Basins.

Information from these SPARROW models can help guide nutrient and sediment reduction strategies throughout the Midwest. Model results provide information on what may be the most appropriate general type of actions to reduce total loading by describing the relative importance of each source, and where to most efficiently place the efforts to reduce loading by describing the distribution of nutrient and sediment loading. By implementing management efforts addressing the major sources of the loads in areas contributing the highest loads, it may be possible to reduce nutrient loading throughout the Mississippi River Basin and thus reduce the size of the hypoxic zone in the Gulf of Mexico; reduce nutrient loading into lakes, and thus reduce the occurrence of harmful algal blooms; and reduce sediment losses, and thus improve the benthic habitat in streams and rivers throughout the Midwest.

Suggested Citation

Robertson, D.M., and Saad, D.A., 2019, Spatially referenced models of streamflow and nitrogen, phosphorus, and suspended-sediment loads in streams of the Midwestern United States: U.S. Geological Survey Scientific Investigations Report 2019–5114, 74 p. including 5 appendixes, https://doi.org/10.3133/sir20195114.

ISSN: 2328-0328 (online)

Study Area

Table of Contents

  • Foreword
  • Abstract
  • Introduction
  • Methods
  • SPARROW Streamflow Model
  • SPARROW Total Nitrogen Model
  • SPARROW Total Phosphorus Model
  • SPARROW Suspended-Sediment Model
  • Model Limitations and Future SPARROW Model Development
  • Summary and Conclusions
  • Acknowledgments
  • References Cited
  • Appendixes 1–5

Additional publication details

Publication type Report
Publication Subtype USGS Numbered Series
Title Spatially referenced models of streamflow and nitrogen, phosphorus, and suspended-sediment loads in streams of the midwestern United States
Series title Scientific Investigations Report
Series number 2019-5114
DOI 10.3133/sir20195114
Year Published 2019
Language English
Publisher U.S. Geological Survey
Publisher location Reston, VA
Contributing office(s) WMA - Earth System Processes Division, Upper Midwest Water Science Center
Description Report: ix, 74 p.; Data Release
Other Geospatial Midwestern United States
Online Only (Y/N) Y
Additional Online Files (Y/N) Y
Google Analytic Metrics Metrics page
Additional metadata about this publication, not found in other parts of the page is in this table