Groundwater Legacies: Quantifying Nitrate Accumulation in the Upper Mississippi River Basin

Nitrate is one of the most common anthropogenic contaminants found in waters of the United States. Increases in nitrate contamination are consistent with the approximately 20-fold increases in chemical N fertilizer application since the mid-20th Century. Rising nitrogen (N) loading rates at the land surface have increased the accumulation of N in surface water, soils, and groundwater. However, the total magnitude of N accumulation in aquifers of the United States remains unknown. Here we provide a first mass estimate of N stored in groundwater within the Upper Mississippi River Basin (UMRB), an area known for intensive row-crop agriculture and a major contributor to Gulf of Mexico hypoxia. Using a novel dataset of nitrate well concentration values, geographic predictors, and a machine learning algorithm, we have produced gridded predictions of nitrate concentrations ranging from the water table surface to 200m below the surface. We find that, in total, a sum on the scale of hundreds of teragrams of N is currently stored in groundwater of the UMRB. This total mass highlights the need to consider long-term groundwater N storage when trying to reduce N loading to surface water and mitigating public health risks from private wells. In addition to providing a starting point for localized studies within the UMRB, this modeling framework demonstrates Random Forest modeling as a powerful tool for estimating contaminant magnitudes on a regional scale.