| Abstract: |
Because all models are simplifications of natural systems it is desirable to use a variety of model constructs for any given natural system. More confidence can be placed in a result, such as the nutrient loading decreases needed to achieve a given reduction in hypoxic area, if this result is supported by multiple modeling approaches. We contribute to this diversity of modeling approaches with an implementation of the Streeter-Phelps river model for the bottom waters of Chesapeake Bay. This model is structurally simple, such that it can be parameterized with a minimum of field data, but retains enough mechanistic detail to allow validation against measured rate processes in the bay. Additional strengths of this model include explicit handling of parameter uncertainty and incorporation of multiple data types in parameterization and forecasting through Bayesian inference and Monte Carlo simulation. Results to date show (1) that Chesapeake Bay hypoxia is strongly influenced by total nitrogen (TN) loading but that the conversion efficiency of nitrogen to hypoxia varies from year to year, (2) specifically that this conversion efficiency has increased since 1980s, (3) that due to this increase in efficiency, TN load reductions of 35%, that previously would have returned the bay to hypoxic areas typical of the 1950s – 1970s, would now be inadequate to return the bay even to the 1980s and 1990s hypoxic areas. |
