Document Type



Virginia Institute of Marine Science

Publication Date



Virginia Institute of Marine Science




Seasonal hypoxia is a characteristic feature of the Chesapeake Bay due to anthropogenic nutrient input from agriculture and urbanization throughout the watershed. Although coordinated management efforts since 1985 have reduced nutrient inputs to the Bay, oxygen concentrations at depth in the summer still frequently fail to meet water quality standards that have been set to protect critical estuarine living resources. To quantify the impact of watershed nitrogen reductions on Bay hypoxia during a recent period including both average discharge and extremely wet years (2016–2019), this study employed both statistical and three-dimensional (3-D) numerical modeling analyses. Numerical model results suggest that if the nitrogen reductions since 1985 had not occurred, annual hypoxic volumes (O2 < 3 mg L−1) would have been ~50–120% greater during the average discharge years of 2016–2017 and ~20–50% greater during the wet years of 2018–2019. The effect was even greater for O2 < 1 mg L−1, where annual volumes would have been ~80–280% greater in 2016–2017 and ~30–100% greater in 2018–2019. These results were supported by statistical analysis of empirical data, though the magnitude of improvement due to nitrogen reductions was greater in the numerical modeling results than in the statistical analysis. This discrepancy is largely accounted for by warming in the Bay that has exacerbated hypoxia and offset roughly 6–34% of the improvement from nitrogen reductions. Although these results may reassure policymakers and stakeholders that their efforts to reduce hypoxia have improved ecosystem health in the Bay, they also indicate that greater reductions are needed to counteract the ever-increasing impacts of climate change.


doi: 10.1016/j.scitotenv.2021.152722


Chesapeake Bay, Hypoxia, Oxygen, Nutrient reductions, Climate change, Water quality management

Creative Commons License

Creative Commons Attribution 4.0 International License
This work is licensed under a Creative Commons Attribution 4.0 International License.

Publication Statement

The model results and data used in this manuscript are permanently available and can be downloaded online from the William & Mary ScholarWorks data repository: st3m-kb83

Frankel et al supplementary (4399 kB)
Supplementary Material