Watershed model: SPARROW


SPARROW (SPAtially Referenced Regression On Watershed attributes) is a watershed model that relates patterns in water quality to human activities and natural processes. Using existing monitoring data, SPARROW analyses the water quality of streams, rivers, and lakes in relation to the location and relative intensity of contaminant sources, landscape characteristics, and environmental factors. This means that SPARROW models can follow the transport of modelled contaminants and nutrients from inland watersheds downstream to larger water bodies, keeping track of their origins and fates.

SPARROW model graphic
Figure 1 - Graphic displaying how the SPARROW model analyzes the impacts of contaminant sources on water quality. Credit: Grabhorn Studios

This can give insight on the causes and effects of challenging and complex environmental issues related to water quality. One such issue found across the transboundary basins is excessive nutrient loading. Human land use practices and activities, like agriculture and wastewater discharge, are compounding the total amounts of nitrogen (TN) and phosphorus (TP) that enter the boundary waters. These are both examples of constituents that SPARROW models can track as they are transported and deposited into receiving lakes or reservoirs.

In excess, these nutrients can become a major problem for water quality. Excessive nutrient loading can cause algal blooms that can be toxic to humans and wildlife, increase the costs of treating drinking water, and limit recreational activities. This nutrient over-enrichment can also lead to eutrophication in downstream waterbodies, which depletes them of oxygen, consequently threatening fish and the overall health of aquatic ecosystems. Examples of lakes that have become eutrophic because of excess binational nutrient loading include: Lake Champlain-Missisquoi Bay (IJC, 2012a), Lake Erie (IJC, 2014a), Lake of the Woods (Clarke and Sellers, 2014) and Lake Winnipeg (Environment Canada and Manitoba Water Stewardship, 2011).

Issues of water quality have broad effects and are very pertinent to the IJC. The Boundary Waters Treaty, which established the IJC in 1909, provides principles for Canada and the United States to follow in using the waters they share. Far ahead of its time, the treaty states that waters shall not be polluted on either side of the boundary to the detriment of health or property on the other side. This applies to boundary water systems as a whole as well as the many IJC Boards including the Red River, the Souris River, and the Rainy-Lake of the Woods boards that have language in their directives to address issues of water quality. Furthermore, since the early 1970s, through the Great Lakes Water Quality Agreement, Canada and the United States have made it a goal to restore and maintain the physical, chemical, and biological integrity of the Great Lakes. To help achieve these goals and address issues of water quality across the transboundary, nutrient reduction strategies are needed that require knowledge of where water quality problems exist, as well as where and from what sources the contributing nutrients originate throughout the watershed. Through applications like SPARROW, modelling helps provide answers to these questions.

Normally, water quality health is determined through water quality monitoring, defined as the sampling and analysis of water and conditions of the waterbody (i.e. a stream, lake, river, or estuary). It evaluates the physical, chemical, and biological characteristics of a water body related to human health, ecological conditions, and designated water uses in the waterbody (US Environmental Protection Agency). Models on the other hand are tools for interpreting such observations. For example, using geographic data models can simulate patterns using both statistical relationships and physical processes represented in the model to develop a more complete picture of water quality issues in a watershed. These findings can be mapped in GIS software. The integration of monitoring and modelling is crucial for our current and future understanding and management of large-scale water quality.

SPARROW modelling results can help:  

  • Determine how to reduce loads of contaminants and design protection strategies;
  • Design strategies to meet regulatory requirements;
  • Predict changes in water quality that might result from management actions; and
  • Identify gaps and priorities in monitoring.

The output of SPARROW models is a spatial representation of total nutrient load and yield, broken down by watershed. In addition, the model can produce a breakdown of the different sources of these nutrients ranging from human activities and land use practices, to environmental sources including agricultural activity, atmospheric deposition, and more.

This data can be explored using Interactive Mappers. Through these mappers, the relationships between contaminant transport, human activities, and natural processes can be visualised through an interactive map and nutrient source charts. They also act as a location to download the model data and results from the SPARROW model.

A map created displaying Total Phosphorous by sub-watershed of the Red-Assiniboine drainage area using mapper data and a screenshot comparing sources of phosphorous in the Red-Assiniboine mapper
Figure 2 – A map created displaying Total Phosphorous by sub-watershed of the Red-Assiniboine drainage area using mapper data and a screenshot comparing sources of phosphorous in the Red-Assiniboine mapper


Several workshops held in 2010 and 2012, with representation from most of the boards, were held to discuss best practices for modelling at the IJC, focusing on hydraulic, hydrologic, and water quality. The SPARROW model was presented as one of the options and a decision was reached to pilot it with the Red-Assiniboine mapper.

After considering many existing water quality models, the IJC decided to use the SPARROW model, a model previously developed by the USGS. This model was selected because it had already undergone extensive peer review, and was appropriate for the scale and purpose of IJC applications. For example, SPARROW models are good at estimating regional nutrient loading and quantifying sources in large basins. In addition, much effort had already gone into the application of SPARROW in the US portion of the Red-Assiniboine basin.

This project was designated as a strategic priority for the IJC under the International Watersheds Initiative (IWI), a 21st century initiative that uses an ecosystem approach, recognizing that environmental systems, in this case watersheds, function as whole entities and should be managed as such, in order to understand the system as a whole. Their reach is not limited to watersheds located next to and straddling the international boundary between Canada and the U.S. With IWI funding, the IJC assembled and supported a strong scientific team to undertake an application of a SPARROW model in the Red-Assiniboine basin. In partnership with the USGS and the National Research Council of Canada (NRCC), and with active participation from several federal, states and provincial agencies, the work began in 2011.

Red-Assiniboine Basin Mapper


Excessive phosphorus (TP) and nitrogen (TN) inputs from the Red–Assiniboine River Basin (RARB) have been linked to eutrophication of Lake Winnipeg. It is important for the management of water resources to understand where and from what sources these nutrients originate, in order to manage these sources and improve the quality of the receiving waters.

At the request of the International Souris River and Red River Boards, the Commission undertook the development and binational application of a SPARROW model for the Red-Assiniboine basin nutrient loading estimation. The RARB straddles the Canada–United States border and includes portions of Saskatchewan and Manitoba in Canada, and North and South Dakota and Minnesota in the United States. This represents the first binational application of SPARROW models to estimate loads and sources of TP and TN by jurisdiction and basin at multiple spatial scales.

The model now has been calibrated and has been consistently applied to the full Red-Assiniboine basin after three years of intensive work that was supported by government partner agencies in both Canada and the U.S. (Jenkinson and Benoy, 2015). This model enables all jurisdictions to better understand water quality dynamics and nutrient loading in this important transboundary basin.

A map made using this model would show the distribution of the total phosphorous or nitrogen deposited into a waterbody by sub-watershed, here the Red and Assiniboine rivers and their tributaries. In theory, a user could see which sub-watersheds are the culprits for pollution. A quick Google Maps search or an analysis of the surrounding environment could explain why the results show up the way they do. Figure 3 shows those areas in the basin that have the highest phosphorus yields and therefore where reduction efforts could be effectively focused. Based on the model, it is estimated that about two-thirds of the phosphorus loading that comes from the Red River into Lake Winnipeg originates in the U.S. portion of the basin. It is becoming increasingly clear that a binational solution is required to address this environmental issue.

A map created displaying Total Phosphorous by subwatershed of the Red-Assiniboine drainage area using mapper data and a screenshot comparing sources of phosphorous in the Red-Assiniboine mapper
Figure 3 – Sample Output from the Red-Assiniboine River SPARROW Model showing Total Phosphorus Yields (kg/km2/yr) by subwatershed across the Basin

The International Red River Board is planning to use the results from the model in support of its basin-wide nutrient management strategy to encourage all impacted jurisdictions to use this information in working towards solutions to help reduce nutrient loading. This modelling work is unique, as it marks the first time that there has been full binational collaboration in the development and application of a common regional water quality model to a transboundary basin in North America.

These results can be further explored with the interactive mapper.

Figure 4 - Screenshot of the mapper created by the USGS for analyzing the Red-Assiniboine basin using SPARROW data

Partners - Acknowledgements

SPARROW Modelling has been a partner driven endeavour.

The major partners were the International Joint Commission (IJC), particularly the International Watersheds Initiative (IWI), and the Wisconsin Water Science Centre at the United States Geological Survey (USGS). Staff at several federal and provincial agencies also contributed their expertise and data to the project: Craig Johnston and Donna Myers (USGS), Richard Burcher and Martin Serrer (National Research Council), Erika Klyszejko and Craig McCrimmon (Environment and Climate Change Canada; Erika now works for the International Joint Commission), Jason Vanrobaeys (Agriculture and Agri-Food Canada), Elaine Page and Justin Shead (Manitoba Conservation and Water Stewardship) and Pam Minifie and Ondiveerapan Thirunavukkarasu (Saskatchewan Water Security Agency). Michael Laitta and Ted Yuzyk from the IJC’s Hydrographic Data Harmonization Task Force set the context for this binational model application. The Red-Assiniboine project was suggested by the IJC’s Red River and International Souris River Boards.

Contact info and references

Building on this success, the IJC is now focusing its effort on the development of a SPARROW model that will cover the Rainy-Lake of the Woods and Great Lakes basins supported by IWI funding. USGS has recently uploaded a mapper for the Great Lakes that focuses on sub-sub-watersheds on the U.S. side.

Graph showing distribution of phosphorus sources in the Great Lakes by lake in tonnes per year (tonnes/yr). Lake Erie has the most phosphorus pollution among any other lake and is the focus of cleanup efforts by IJC boards and provincial and state agencies.
Figure 5 – Graph showing distribution of phosphorus sources in the Great Lakes by lake in tonnes per year (tonnes/yr). Lake Erie has the most phosphorus pollution among any other lake and is the focus of cleanup efforts by IJC boards and provincial and state agencies.

Learn More

USGS Page including FAQ: http://wi.water.usgs.gov/nutrients/sparrow


For more info please contact:


Michael Laitta, laittam@washington.ijc.org

Lizhu Wang, lizhu.wang@windsor.ijc.org

Wayne Jenkinson, jenkinsonw@ottawa.ijc.org


Dale M. Robertson, dzrobert@usgs.gov

David Saad, dasaad@usgs.gov


Ivana Vouk, ivana.Vouk@nrc-cnrc.gc.ca


Benoy, G., Jenkinson, R., Robertson, D., & Saad, D. (2016). Nutrient delivery to Lake Winnipeg from the Red—Assiniboine River Basin – A binational application of the SPARROW model. Canadian Water Resources Journal / Revue Canadienne Des Ressources Hydriques, 41(3), 429-447. Retrieved from https://www.tandfonline.com/doi/full/10.1080/07011784.2016.1178601

Jenkinson, R.W., and Benoy, G.A. (2015). Red-Assiniboine Basin SPARROW Model; Development Technical Document. National Research Council of Canada. Ottawa, ON. 65pp. Retrieved from https://ijc.org/sites/default/files/IJC%20RA%20Model%20Development%20Report%202015%20FINAL.pdf

“Overview of Watershed Management.” (n.d.) Watershed Academy Web: Distance Learning Modules on Watershed Management. U.S. Environmental Protection Agency. Retrieved from https://cfpub.epa.gov/watertrain/pdf/modules/monitoring.pdf.