Modeling the spread of radioactivity in seawater

When earthquake-triggered tsunami waves hit Japan in March, the surging water overtopped seawalls and caused massive damage. On top of the loss of life and general destruction, the disaster resulted in a release of radioactive seawater. Researchers Changsheng Chen of the University of Massachusetts at Dartmouth and Robert Beardsley of the Woods Hole Oceanographic Institute are investigating where that radioactivity may spread, using a mathematical model they developed in part through funding by the MIT Sea Grant College Program.

The meltdown at the Fukushima Daiichi Nuclear Power Plant released several different types of radioactive particles, but scientist are most interested in tracking a particularly persistent radioactive particle known as cesium-137. This chemical has a half-life of 30 years — in other words, it takes about three decades for the particle’s radioactivity to fall by half. If ingested, cesium-137 accumulates in living tissue, and it can travel up the food chain if one animal consumes a plant or animal that ingested contaminated particles.

Early estimates by oceanographers Chen and Beardsley suggest that the radioactive particles would disperse throughout the ocean differently at different depths. In some cases, contaminated seawater would reach the western coast of the United States in as little as five years. The researchers have been building and refining their mathematical model of ocean circulation since 2003, and they recently received funding from the National Science Foundation to do an in-depth, collaborative study of how the particles would move through the ocean.

The mathematical model that Chen and Beardsley created and use, known as the Finite Volume Community Ocean Model, or FVCOM, has been in existence for about a decade. Unlike other ocean models, this program can be used to examine what happens to different parts of the environment at different levels of detail. In other words, the researchers can use very small grid cells to compute information about what’s going on in the complex near-shore ocean environment, and at the same time they can use big grid cells to simulate, in broad strokes, the calmer currents of the open ocean.

That’s particularly important for the nuclear-particle tracking study, in which the scientists need to be able to make detailed forecasts of the coastal areas and also need to run the simulation quickly — something that wouldn’t be possible if they had to use the same level of detail across the whole ocean simulation.

The unique variable-grid-size method employed by FVCOM is an ideal tool for other projects as well. Chen and Beardsley are also using FVCOM to conduct modeling experiments to look at the impact of long- and short-term processes on water quality in Massachusetts Bay, as part of an MIT Sea Grant-funded project. This may help scientists understand how those factors impact the occurrence of red tides — toxic algal blooms that decrease dissolved oxygen and disrupt shellfish fisheries.

Creating a mathematical model of a small coastal area can be tricky, Chen explains, because the ocean is undivided. Changes in one part of the ocean can alter conditions everywhere else. But scientists study small, well-defined areas. For example, scientists who want to study the Gulf of Maine might be tempted to draw a straight line across the open ocean between Cape Cod, Mass., and Cape Sable in Nova Scotia. But that line doesn’t even come close to reality. Water is always flowing.

Chen and Beardsley have been working to find new ways of estimating “boundary conditions” — the information that the model uses to simulate ocean water flowing into and out of the region that scientists are studying (in this example, coastal New England). One long-term solution they’re considering is a fast and efficient global version of FVCOM, which would use a variable grid to calculate how what’s happening in the ocean near Maui impacts what happens off the coast of Maine.

“It’s a really nice evolution, driven primarily by scientific questions,” Chen says, commenting on how FVCOM has grown from a relatively little-known estuarine circulation model into the powerful, open-source ocean model that can generate predictions of both international and local importance.