Teams of scientists have begun working with Bay Program officials to review and improve the model used to predict how the Chesapeake will respond to nutrient and sediment reductions.

The goal is to have an improved water quality model, recently the subject of a critical review, ready early next year as a tool to help state and federal officials decide the magnitude of nutrient and sediment reductions needed to achieve a “clean” Bay by 2010.

An academic review team selected by the Bay Program’s Scientific and Technical Advisory Committee issued a report late last year that found fault with several key parts of the model, which is considered to be one of the most sophisticated and complex of its type in the world.

But the review team concluded that the model “does not currently provide information suitable for major management decisions and that use of the model for such purposes should be suspended.” [See “After review, Bay Program moves to improve water quality model,” January-February 2000.]

The team said the model failed to accurately simulate several key processes in the Bay, including the amount of algae that should be produced by nutrients.

The model review report was formally presented to the Bay Program’s Implementation Committee in February.

In a memo to the committee, STAC said the report should be “embraced as a wake-up call for the the technical and management communities broadly involved in the Bay Program to work in a concerted effort to improve the model.”

But STAC acknowledged that it “does not believe” the use of the model was inappropriate in setting nutrient reduction goals in Virginia tributaries — the only decision-making application of the model to date.

“The model was useful in demonstrating to decision-makers the linkages between nutrient inputs, water quality and living resources and the relative effects of point source controls among those tributaries,” the memo said. “This was helpful in determining where to target efforts.”

Further, it said, general conclusions from the model about the impact of nutrient reductions in the tributaries were “highly unlikely to change with model refinement” although predictions might get slightly more accurate about exactly how much improvement to expect.

Still, the STAC memo said, the model has “significant problems” which are “not scientific fine points.”

Assertions that the water quality model has serious flaws have been sharply disputed by many. The Chesapeake Bay Modeling Team, a group of the U.S. Army Corps of Engineers and EPA scientists who oversee model development, said in a written response that the goal of the modeling effort was to “reduce uncertainty” in decision-making, not perfection.

The modeling team wrote that “if perfection needs to be the criterion, then no calculation will measure up.” But they said “shortcomings [in model predictions] are understood and can be considered when interpreting results.”

Likewise, Virginia members of the Bay Program’s Modeling Subcommittee wrote that the model had been effective in developing goals for the York, Rappahannock and James rivers by showing how changes in land-based activities would affect water quality. They wrote that there was “no other tool that we can use to accomplish this difficult task.”

While the model has shortcomings, the Virginia officials wrote, they do not warrant the conclusion that its use should be suspended.

The Implementation Committee tended to agree. Officials acknowledged that no model could hope to perfectly represent a system as complex as the Bay, but insisted that decisions made with the help of the model would be better than decisions made without it.

At the same time, they supported efforts to improve the model. The EPA’s Bay Program Office has already assembled teams of scientists who are reviewing and helping to improve key physical and biological processes in the model.

Also, the Bay Program has been putting documentation materials for the model — information about how it works and peer-reviewed articles about its operation — on its web site. The academic review team had criticized the lack of availability of such information, making it impossible for outsiders to understand how the model works. In addition, the Bay Program is ultimately planning to make all of its computer models available on its web site.

The academic review team had faulted the Bay Program for never clearly establishing what is expected from its models. In response, the Implementation Committee is also assembling a group of scientists and managers that will more clearly describe the specific management objectives and intended uses of the model, and to develop a document that more clearly states the model’s strengths and weaknesses as an aid to people who interpret model results.

Bay Program officials anticipate that refinements to the water quality model will be completed by the end of the year so it can be used to help determine the amount of nutrient and sediment reductions needed to clean up the Chesapeake.

The draft Chesapeake 2000 Bay Agreement calls for cleaning up the estuary by 2010. If that deadline is not met, the EPA, under a court ruling last year, would require the development of a cleanup plan known as a Total Maximum Daily Load for the watershed, which would force far more prescriptive, and potentially costly, cleanup measures.

To meet that timetable, estimates of new nutrient and sediment reduction goals are needed next year. The water quality model uses information from the Bay Program’s watershed and airshed models — which estimate the amount of nutrients and sediments that would enter the Chesapeake under different control strategies — and predicts how the Bay would respond.

Even without using the Bay water quality model, scientists and managers increasingly have recognized that far more nutrient reductions were needed to restore the Bay than had previously been planned.