Back to Woods Hole Oceanographic Institution's Homepage
  • Connect with WHOI:
true expositive en-US Lonny Lippsett WHOI

Building a Computer Model to Forecast Red Tides

A variety of variables goes into the mathematical mix that simulates harmful algal blooms

SHARE THIS:   TOOLS: print this page

The algae Alexandrium fundyense are notorious for producing a toxin that accumulates in shellfish such as clams, mussels, and oysters, leading to paralytic shellfish poisoning in humans. The microscopic plants are naturally distributed in New England waters, but in some years—2005, for example—a confluence of factors causes major blooms that spread down the coast and force officials to close shellfish beds.

To protect consumers, and the livelihoods of shellfishermen from Maine to Massachusetts, researchers would like to able to forecast when and where Alexandrium will bloom and to explain how and why. In a project sponsored by the Woods Hole Center for Oceans and Human Health, scientists Dennis McGillicuddy and Ruoying He of Woods Hole Oceanographic Institution have developed a mathematical model of Alexandrium dynamics—a series of equations that captures the physical and biological factors involved in harmful algal blooms in New England.

Millions and millions of calculations

Working with colleagues in the biology lab of WHOI Senior Scientist Don Anderson, McGillicuddy and He entered a range of factors into their model: the speeds and directions of ocean currents, water temperature and salinity, winds, surface heat exchanges, tides, river runoff, and the distribution and behavior of Alexandrium cells in the water and in seafloor sediments. The team—aided by WHOI researchers Valery Kosnyrev, Olga Kosnyreva, and Larry Anderson—also incorporated other computer models that describe waterborne nutrients, solar radiation, and large-scale motions in the North Atlantic Ocean.

“The beauty of a numerical model,” McGillicuddy said, “is that it can be used to investigate extraordinarily complex systems by posing the problem in terms of a few guiding principles expressed as mathematical equations.”

Plug numbers into the equations, and the model develops a big picture of how the algae spread as a result of the interactions of all those variables. Using a 32-processor “mini-supercomputer,” as He calls it, it can take two days to make all of the calculations necessary to simulate three months of Alexandrium dynamics at more than a million different points in the Gulf of Maine.

How does a bloom end?

The project has been under way for several years, but the work grew more intense and fruitful in 2005. A devastating bloom provoked researchers from WHOI and a dozen federal and state agencies to launch a substantial effort to gather observations of ocean conditions during the event. That fieldwork has allowed He and McGillicuddy to run an extensive “hindcast” of the 2005 bloom; that is, they were able to plug real environmental data into their computer model and see if it could reproduce what actually happened.

“We wanted to compare the model against new observations to see if we really understood the system or if we were just fooling ourselves,” said McGillicuddy.

The model successfully captured how intense spring northeasterly winds and prodigious river runoff, coupled with a huge seed population of Alexandrium cysts, blossomed into a red tide that blanketed New England waters from Canada’s Bay of Fundy to Martha’s Vineyard. The event had an estimated $20 million impact on the seafood industry in New England, according to researchers at the WHOI Marine Policy Center.

The principal flaw in the model’s 2005 projection is that it could not replicate the end of the bloom, which occurred in July. “What processes regulate the end of a bloom and make it shut down?” said He. “We just don’t know enough yet about mortality in Alexandrium, and we are still trying to improve the modeling of this process.”

But the team was encouraged enough to try to track the 2006 season as it develops. Though they are not willing to make public predictions, He and McGillicuddy have been plugging in field observations this spring, letting their model run into the future, and then seeing how well it did. So far, their computerized bloom seems to jibe with the milder blooming realities of 2006.

This research is supported NOAA's Center for Sponsored Coastal Ocean Research and by the National Science Foundation and the National Institute of Environmental Health Sciences through the Woods Hole Center for Oceans and Human Health.


Dennis McGillicuddy (left) and Ruoying He deploy a drifter to track the flow of currents around the Bay of Fundy. (Photo by Mike Carlowicz, WHOI)

Dennis McGillicuddy and Ruoying He have developed a computer model of the dynamics and dispersion of the harmful algae Alexandrium fundyense. (Photo by Mike Carlowicz, WHOI)



"Hindcast" model of 2005 red tide

Oceanographers have developed a computer model to simulate last year's historic bloom of toxic algae.


Model of 2006 Alexandrium season

WHOI researchers have been running their model in real time this spring.

Woods Hole Oceanographic Institution is the world's leading non-profit oceanographic research organization. Our mission is to explore and understand the ocean and to educate scientists, students, decision-makers, and the public.
© Woods Hole Oceanographic Institution. Online edition: ISSN 1559-1263. All rights reserved