Skip to content

Connect with WHOI:

Corralling the Wild and Wooly Southern Ocean For his Ph.D. work in the MIT/WHOI Joint Program, Matt Mazloff merged sparse observations of the Southern Ocean with a state-of-the-art ocean circulation model to produce estimates of ocean conditions of greatly increased accuracy. The model can fill in the blanks to describe what’s going on in places where no observations have been taken. This map shows the speed of the clockwise Antarctic Circumpolar Current on May 12, 2006, increasing from slow-moving water (blue) to water moving more than one mile per hour (dark red). This research was supported by the National Ocean Partnership Program.  (Image courtesy of Matthew Mazloff, Source: San Diego Supercomputer Center, UC San Diego)

Corralling the Wild and Wooly Southern Ocean

Graduate student creates supercomputer model to tame a vast, remote ocean


Matt Mazloff fishes out a postcard. It’s a simulated aerial view of the bottom of the world, with Antarctica in the middle and the tips of South America, Australia, and a smidgen of Africa peeking in from the corners. But it’s not the land that Mazloff is interested in. It’s the huge expanse of water encircling Antarctica—the Southern Ocean.

“I’m just fascinated by it, there are so many open questions,” he said.

Mazloff has fed that fascination as a graduate student in the MIT/WHOI Joint Program, creating a computer model that simulates the complex dynamics of the world’s southern oceans—a wild and woolly region that is roughly 13 times the size of the United States and very poorly understood. He has developed the so-called Southern Ocean State Estimate, or SOSE—a project that involves using reams of measurements and observations—taken by ships, orbiting satellites, drifting floats, and even electronic transmitters strapped to elephant seals—and synthesizing them in accordance with the laws of physics and fluid dynamics.

Mazloff, a physical oceanographer, feeds it all into a supercomputer at San Diego Supercomputing Center.  The computations create simulations—resembling those that meteorologists have for atmospheric weather—of how the ocean is moving and working. The model can fill in the blanks in places where no observations have been taken.

“What’s created, then, is a model that accurately tells the story of the Southern Ocean’s state at any location and over a given period in time,” he said.

From micro- to macrocosm

For oceanographic purposes, the Southern Ocean extends from the coast of Antarctica north to about 35 degrees southern latitude (for comparison with the Northern Hemisphere, this is analogous to the area north of Charlotte, North Carolina) and connects the Atlantic, Indian, and Pacific Oceans, which radiate from it like spokes on a wheel. The vast Southern Ocean stores and moves huge amounts of heat and the greenhouse gas carbon dioxide around the globe, and so it plays a critical role in regulating Earth’s climate. It is also one of the world’s most fertile oceans, teeming with microscopic marine plants and animals that feed a wealth of fish, penguins, seals, whales, and seabirds.

Despite its importance, the Southern Ocean has defied study mainly because of its remoteness, cold, windy weather, high-rise tidal surges, ice-choked waters, and complete darkness for nearly half the year. Only within the last decade, with technological advances, have more data come in from the Earth’s southern polar region.

The data, though extensive, still “has a ton of holes in it,” Mazloff said. “So, we’re trying to use the laws of physics, as we can best represent them, to fill all the gaps in.”

Laboring over physics was not what the 30-year-old Mazloff envisioned himself doing. Born in Northampton, Mass., he was accepted into the engineering department at the University of Vermont. He couldn’t decide on which area he wanted to study, “so they said, ‘Why not take physics?’ I never ended up going back to engineering.”

He earned a master’s degree in biophysics, specializing in the interaction of DNA and lipids. But he was frustrated by the work, because “everything was so small,” he said. Then, in 2001, he heard from a friend who was living on Cape Cod about the MIT/WHOI Joint Program in oceanography. Intrigued and yearning for something different, Mazloff applied.

“I became a student partly so I could live on the Cape,” he said with a broad smile.

Trillions times billions of data points

At WHOI, Mazloff applied his training in physics, studying how the ocean flows at topographical (land) edges and building a miniaturized version of the Labrador Sea to study the geophysical fluid dynamics in detail. It was “clean, idealized” work, he said.

Soon after he sought out his Ph.D. advisor, Carl Wunsch, and became part of the Estimating the Circulation & Climate of the Ocean (ECCO) consortium, a group of researchers seeking to model the oceans and their climate impacts globally.

He was especially drawn to the Southern Ocean and the chance to help create a model that would yield the most accurate, comprehensive picture to date of how this big mysterious region works.

“It’s quite remarkable for a student to take on such a complicated problem,” said Wunsch, a physical oceanography professor at MIT. “He convinced me to do it.”

Mazloff’s model state consists of more than 10 trillion terms constrained by more than one billion ocean observations. “It’s like trying to draw a best-fit line through a billion points,” Mazloff said as he looked at rows of numbers cascading on his computer screen. “That’s why you need a supercomputer.”

Culminating the research for his Ph.D. degree, Mazloff produced and analyzed model simulations spanning 2005 and 2006 that revealed Southern Ocean dynamics with increasing accuracy and unprecedented detail. He defended his dissertation in July 2008, and two weeks later drove cross-country to continue his ocean studies as a postdoctoral researcher at the Scripps Institution of Oceanography.

“The MIT/WHOI Joint program has allowed me to make a practical contribution to climate research,” he said, “and I am very grateful for that.”

Mazloff’s research as a graduate student was supported by grants and contracts from the National Aeronautics and Space Administration and the National Ocean Partnership Program. Computational resources were provided by the San Diego Supercomputer Center under a National Science Foundation Teragrid grant. As a student, he was supported by The J. Seward Johnson Fund and The Virginia Walker Smith Fund.

Featured Researchers