SPARROW nutrient modeling: Mississippi/Atchafalaya River Basin (MARB)

SPARROW (SPAtially Referenced Regression On Watershed attributes) models track the transport of water, nutrients (particularly nitrogen and phosphorus), and sediment from local inland watersheds to regional, coastal waters by explaining spatial patterns in
stream water-quality conditions in relation to human activities and natural processes.

2012 Mississippi/Atchafalaya River Basin (MARB) SPARROW Models

To further refine the estimated sources of nutrients to the Gulf of Mexico, SPARROW models were developed using nutrient input data from management practices similar to 2012, and developed using smaller catchments (average of 2-3 square kilometer catchments) to enable improved spatial descriptions of nutrient sources. The models were calibrated using total phosphorus and total nitrogen loads from sites throughout the Midwest. These models are described in detail in U.S. Geological Survey Scientific Investigation Report 2019-511. Results from the updated and refined models were applied to the Mississippi/Atchafalaya River Basin, which was published in the Journal of the American Water Resources Association in 2021.

Using the SPARROW (SPAtially Referenced Regression On Watershed attributes) model, the study determined that catchments in the Corn Belt and along the Mississippi River had the highest yields (mass per unit area) of N and P. Agricultural activities were the largest source of N and P, but local inputs from wastewater treatment plants in some cases also were large contributors, and natural sources were sizable contributors of P. Results of the model were again used to rank each of the states in the Basin based on their relative nutrient contributions to the Gulf of Mexico.

An associated online mapping tool can be used by water-quality mangers to identify where the largest sources of nutrients are throughout the Mississippi/Atchafalaya River Basin, and identify the types of major nutrient sources contributing to the loads in the various areas, which can be used to guide actions to reduce nutrient loading to the Gulf of Mexico. The updated data and finer spatial resolution will better inform these management actions.

2002 Mississippi/Atchafalaya River Basin (MARB) SPARROW Models

Sources/Usage: Some content may have restrictions. View Media Details

To refine estimated sources of nutrients to the Gulf of Mexico, SPARROW models were developed specifically for the MARB using nutrient input information from 2002 and refined environmental-setting information. Results of these models were published in the Journal of Environmental Quality in 2013 and Journal of the American Water Resources Association in 2013. The refined models enabled USGS researchers to determine the importance of each nutrient source, particularly specific agricultural sources, more accurately than with previous models. Nutrient sources throughout the MARB are described in detail in the Journal of Environmental Quality in 2013.

These models demonstrated that:

• Highest nitrogen yields were from the Corn Belt (centered over Iowa and Indiana)
• Highest phosphorus yields were scattered throughout the MARB.
• Agricultural input (fertilizer, manure, and fixation) was the dominant source of nitrogen and phosphorus
• The 2002 refined SPARROW models showed that unconfined manure was a less important source of phosphorus than previous SPARROW models showed
• Urban sources (wastewater treatment plants and urban nonpoint) were more important than found with previous SPARROW models.

The newly estimated nitrogen and phosphorus yields throughout the MARB were used to re-rank the various HUC8s, states, and subbasins based on their relative nutrient contributions to the Gulf (described in detail in the Journal of the American Water Resources Association in 2014). Relatively small changes in area rankings (compared to previous models) were found for nitrogen, whereas larger changes were found for phosphorus. Spatial differences in yields from previous studies resulted from different descriptions of the dominant sources (N yields are highest with crop-oriented agriculture and P yields are highest with crop and animal agriculture and major WWTPs) and different descriptions of downstream transport. This information may help managers decide where efforts could have the largest effects (highest ranked areas) and thus reduce hypoxia in the Gulf of Mexico.

1992 Nation-scale SPARROW Models applied to the Mississippi/Atchafalaya River Basin ​​​​​​​

The 1992 Mississippi/Atchafalaya River Basin (MARB) portion of national-scale SPARROW models were primarily based on 1992 data and published in Environmental Science and Technology in 2008. Results from these models were used to evaluate watersheds throughout the MARB that deliver nitrogen and phosphorus to the Gulf of Mexico.
The 1992 MARB portion of the national SPARROW models showed:

• While Illinois, Iowa, Indiana, Missouri, Arkansas, Kentucky, Tennessee, Ohio, and Mississippi make up only one-third of the 30-state MARB area, they contribute more than 75% of nitrogen and phosphorus to the Gulf.
• Agricultural nonpoint sources contributed more than 70% of the nitrogen and phosphorus delivered to the Gulf, versus only about 9 to 12% from urban sources.
• Noncaptured animal manure on pasture and rangelands contribute nearly as much phosphorus as cultivated crops, 37% versus 43%, suggesting that the wastes of unconfined animals is a much larger source of phosphorus in the MARB than previously recognized.
• Atmospheric contributions also were important, accounting for 16% of nitrogen.

• Corn and soybean cultivation was the largest contributor of nitrogen to the Gulf. The study reported that 66% of nitrogen originated from cultivated crops, mostly corn and soybean crops, with animal grazing and manure contributing only about 5%.

Noncaptured animal manure on pasture and rangelands contribute nearly as much phosphorus as cultivated crops, 37% versus 43%, suggesting that the wastes of unconfined animals is a much larger source of phosphorus in the MARB than previously recognized.Read the full press release on the MARB SPARROW model results.
Another goal of the 1992 national-scale SPARROW models was to develop a probabilistic ranking of individual watershed nutrient contributions. To do this, uncertainties in the estimates of nutrient yields from the models were incorporated into a statistical ranking procedure to determine the probability that a watershed was within the top specified number of watersheds delivering the highest nutrient yields to the Gulf. This procedure was published in the Journal of the American Water Resources Association in 2009.
This statistical procedure showed that 11 HUC8 watersheds could be reliably ranked in the top 150 category for most total nitrogen (3 for most total phosphorus) with 90% certainty. Although only 11 watersheds could be confidently placed into the top 150 HUC8 category, numerous watersheds could be removed from consideration of being in the top 150 category because of model uncertainty. A total of 513 HUC8 watersheds for total nitrogen and 505 watersheds for total phosphorus could be reliably placed outside of the top 150 HUC8 category with 90% uncertainty. Probabilistic ranking of watershed nutrient yields can assist water managers, policy makers, and scientists in identifying watersheds that may be primarily responsible for nutrient delivery to the Gulf of Mexico.

Read the full press release on the MARB probabilistic ranking.