EPA picks VIMS computer model in $1.6 million deal to guide Chesapeake Bay pollution decisions

A new computer model from the Virginia Institute of Marine Science was recently chosen to replace an aging system used to guide the massive Chesapeake Bay restoration project.

The U.S. Environmental Protection Agency awarded the institute under William & Mary $1.6 million for SCHISM: the Semi-implicit Cross-scale Hydroscience Integrated System Model. The Chesapeake Bay Program has been using a computer model first developed in the 1990s to forecast conditions in waterways and steer population targets.

Officials said the more granular and sophisticated system will better capture the state of the bay and inform decisions about how to manage it.

The ‘90s model, which has been revamped over the years, was innovative in its time. But officials became increasingly aware of the need for a version that would be “state of the science,” said Lew Linker, modeling coordinator for the EPA’s Chesapeake Bay Program office in Annapolis. The bay program is a large coalition that started in the 1980s and includes nonprofits, academic institutions, and local, state and federal governments.

Computing power has drastically increased, of course. But climate change is another complicating factor, Linker said.

Though the program has made progress toward reducing excess nutrients and sediment that pollute the bay, the goalposts keep shifting because of warming waters, sea level rise, impacts to underwater grasses and more.

“We haven’t had the influence we thought we would have, and it turns out that climate change is that headwind,” he said.

SCHISM aims to help that, by including more detail. The previous model stops at the current shoreline, said Joseph Zhang, a research professor at VIMS’ Center for Coastal Resources Management who spearheaded SCHISM. The shoreline is projected to move landward in the coming decades as sea levels rise, which his model incorporates.

The program’s development goes back nearly two decades. At the time, Zhang worked at a graduate institute in Oregon. He started working on the technology because he saw an urgent need for a robust and hyperlocal system.

He arrived at VIMS several years later and refined it, incorporating conversations with local and state officials.

The EPA put out a request early last year for a fine-scale model of the Chesapeake Bay and chose SCHISM after a selection process.

Visualizing an advanced computer model is tough. Zhang said to think of it somewhat like how a weather forecast uses a variety of factors.

SCHISM is like a forecast for the ocean. Rather than whether to expect rain or wind, it helps understand what conditions by simulating water motion and other elements.

The higher resolution is similar to going from a blurry image of a map to one that’s crystal clear and fully zoomed in.

Another advantage is its ability to be merged with other models, said Kirk Havens, director of the coastal resources center at VIMS.

Leaders in Hampton Roads could combine it with a system focused on groundwater, for example, to assess the impact of sea level rise on drinking water infrastructure.

The central use of SCHISM will be to figure out how restoration actions influence bay pollution.

Restoration officials use a so-called “pollution diet” to limit nutrients and sediment. Created in 2010, the Total Maximum Daily Load is a planning tool from the EPA that calculates the maximum amount of a pollutant that a body of water can receive and still meet water quality standards.

Zhang’s computer model can be used to test whether say, a wetlands project at the Hague to achieve those standards will work as intended.

It also provides a quicker time turnaround to look at different scenarios, Havens said.

That turnaround: a decade-long projection calculated in a single day. A locality can input what it wants to measure, run it for a day and learn what will happen in a very specific watershed over the next 10 years.

The EPA agreement is for six years, the first three of which Zhang’s team will use to work with federal officials to tailor the model for the agency’s needs. They’ll work out the remaining kinks for the three years after a 2025 launch.

Linker said modeling is a core component to the bay program’s operations, working in tandem with live research and monitoring.

“We’re going to understand the bay on a finer scale, in the shallower waters which is really where most of us interact,” he said. “It’s going to be represented and simulated better than ever before, and help everyone in the watershed understand what our shared responsibilities are.”

Katherine Hafner, 757-222-5208, katherine.hafner@pilotonline.com