Published online 28 June 2010 | Nature | doi:10.1038/news.2010.280

News: Q&A

Scratching the subsurface

The Deepwater Horizon oil spill puts ocean-current modelling to the test.

Ruoying He, whose ocean circulation model is being tested to the limit by the BP spill.Ruoying He

Every morning at 2 a.m., a high-performance computer cluster in Raleigh, North Carolina, named henry2 dips into three servers and fishes out a few hundred megabytes of data on two vast areas of water — the Gulf of Mexico and a swathe of ocean off the southeastern United States known as the South Atlantic Bight.

The data include ocean temperature and surface wind, among other variables. Henry2 feeds them into a carefully crafted model that simulates how ocean waters in the region will circulate over the next three days.

For more than two years, this routine operated with little fanfare. Then, on 20 April, the Deepwater Horizon drilling rig exploded. Suddenly, as a stretch of US coastline was revealed as a potential disaster zone, henry2's daily forecast mattered in a new way. Oceanographer Ruoying He, who heads the Ocean Observing and Modeling Group at North Carolina State University in Raleigh, talks to Nature about how the spill tests the limits of his model.

How did things change for you after 20 April?

I feel there's definitely more relevance for my model now. The South Atlantic Bight is in my backyard, only three hours away by car, and in order to make predictions we need to understand circulation in the Gulf of Mexico; it was just a coincidence that the spill happened in that area. If it hadn't, we would have just continued doing circulation simulation.

After the community realized the seriousness of the event, we added the location of spilled oil to the simulation. My group has collaborated with Robert Weisberg and his colleagues at the University of South Florida in St Petersburg since April, examining satellite images of the spill to figure out where the oil is.

On the basis of that location, we deploy numerical particles that trace the oil in the simulations, and then we watch how the currents drive the particles around.

For much of the time, scientists just do things they like to do. Then events like this happen, and we feel that science can contribute more to practical societal problems. Our group is one of several teams running ocean circulation forecast models used by the US National Oceanic and Atmospheric Administration to generate daily predictions of the oil's movement.

Now I check our modelling website on my iTouch while I eat breakfast.

How accurate is your model?

We validate predictions using physical data from satellite and oceanographic sensors. Our predictions are consistently simulating the observations. We're confident that the big picture is on track, but our model is limited by the spatial resolution of 5 kilometres. It can't pinpoint where the oil is going up the beach. Models with better resolution exist, but they're still in research mode.

What do you need to make your predictions more accurate?

Oil trajectory simulations based on circulation models so far predict movement only on the ocean surface, but such information might be misleading. And there is still no consensus on the volume of oil spilled.

No one really knows how the oil is distributed in deep waters or how well our models perform at that level. But there's an urgency to quantify that. My group has just received National Science Foundation Rapid Response Research funding to carry out a three-dimensional simulation of the horizontal and vertical distribution of oil plumes.

We need data on the subsurface oil distribution to constrain and validate our model at depth. Researchers have been deploying instruments into the deep waters since last month, using our models to guide their mission. It's a two-way interaction.

All research activities are in rapid response mode, so not all in situ observations are available to us at this point. Still, having some data is much better than having none.

What is happening in the Gulf right now?

An eddy has separated from the northernmost portion of the Loop Current and stands between the Deepwater Horizon well and the Loop Current itself. So most of the spilled oil has not been entrained in this strong current.

So, what do you think will happen in the future?

The eddy may reattach and detach several times, and will eventually drift westward in the Gulf. When that happens, the Loop Current may eventually re-extend into the northeastern Gulf, and any oil that is entrained in it will eventually pass off the shores of Florida, Georgia and the Carolinas. Around Cape Hatteras, the stream turns east and heads for Europe.


Oil is degraded by sunlight and consumed by microorganisms. If it stays contained in the ocean without touching any beaches, it will gradually dissipate. But because the Loop Current flows right against the Florida Keys, there is the fear of oil getting onto South Florida beaches. My best guess is that North Carolina will see some forms of oil, probably tarballs, in the next several months.

There's another concern: hurricane season has already started. Hurricanes come with strong mixing, moving the deep ocean oil plume to the surface. Newly surfaced oil could be pushed onshore by storm surges.

In that sense, hurricanes may not help the oil relief effort. We didn't expect to see such a nasty thing happening. 


If you find something abusive or inappropriate or which does not otherwise comply with our Terms or Community Guidelines, please select the relevant 'Report this comment' link.

Comments on this thread are vetted after posting.

  • #61203

    I think of myself a libertarian. But I think my feelings only extend towards individual rights. Corporations are amoral creatures and left by themselves their lack of morals or consideration of right and wrong is disturbing and needs to be mitigated.

Commenting is now closed.