August 2, 2010 | When Synte Peacock (NESL/CGD) helped design a computer modeling study to simulate the movement of oil from the massive spill in the Gulf of Mexico, she thought reporters might be interested. (The study technically simulates dye in place of oil, which undergoes a multitude of biological, chemical, and physical changes as it disperses.) But she had no idea that the resulting simulations would be featured in the national media and attract hundreds of thousands of viewers on YouTube.
A team in CISL’s Visualization and Enabling Technologies Section—Tim Scheitlin, Mary Haley, Michael Burek, Rick Brownrigg, and Don Middleton—helped create the animations, working with Synte and her two collaborators (Mathew Maltrud from Los Alamos National Laboratory and Martin Visbeck from IFM-GEOMAR at Germany’s University of Kiel).
The research, presented in a series of animations, shows that a substance released at the spill site would likely get caught in the Gulf's powerful Loop Current at some stage over the following months, and then flow around the southern tip of Florida and northward along the Atlantic coast. The simulations were not able to address whether the oil was likely to affect coastal areas or what form it would take at sites far from the northern Gulf.
The NCAR & UCAR Media Office (housed in UCAR Communications) issued a news release on June 3 and distributed the animation. With the public deeply concerned about the spill, reporters began calling almost immediately.
The story was picked up by nearly 2,000 print, broadcast, and online media outlets around the nation and world, including the New York Times, Wall Street Journal, “CBS Evening News with Katie Couric,” “Good Morning America,” and “The Today Show.” The main animation was also posted to UCAR/NCAR’s YouTube channel on June 3, where it had accumulated more than 750,000 hits three weeks later. A similar animation using Google Earth software posted on June 17 claimed several thousand hits within one week.
For Synte, the experience was a crash course in the media. The training she received from the Media Office paid off when she handled four television interviews and about a dozen interviews with newspaper and radio reporters on the first day alone. The interview requests haven’t stopped since, and CGD director Bill Large has stepped in to help field some of them.
A still image from one of 20 simulations showing the possible trajectory of dye 130 days after the spill. (See the full animation here.) The research team’s original modeling indicated that, at some point in the six months following the spill, the Loop Current was likely to intercept the oil and transport it toward the Atlantic coast. Natural instabilities in the Loop Current, however, led it to spin off a large eddy that prevented the oil from entering the current. Modeling done by the team in late July indicated that some of the oil/dispersant mixture would likely enter the Atlantic in the next six months, with the timing dependent on when the Loop Current reestablished itself. The researchers expect the mixture to be mixed in the upper water column of the Atlantic ocean, rather than a surface slick.
“It’s been an interesting experience and it raises a larger question of whether we as scientists should try to do more things which are relevant to society,” Synte says. “We have the tools and expertise here to apply our models to things that the general public is interested in and which have important political and societal implications. Clearly there are a lot of people at NCAR who are pursuing research along these lines, but should we be doing even more?”
The animations generated even more media coverage of NCAR than the 2007 IPCC assessment, when the Media Office organized a pair of national teleconferences for reporters. “It was quite a busy couple of weeks,” says David Hosansky, head of media relations. “I’d be calling one reporter back about Synte’s work and my phone would light up with two incoming calls. But it was great to see all the media interest.”
To carry out the research, Synte and her co-authors used the Parallel Ocean Program, a component of the CCSM, conducting the simulations on supercomputers at the New Mexico Computer Applications Center and Oak Ridge National Laboratory. Specifically, they ran simulations to see how a liquid released at the spill site would disperse and circulate. They tracked the rate of dispersal in the top 65 feet of the water and at four additional depths, the lowest just above the seabed.
The simulations indicated that a liquid released at the ocean’s surface at the spill site was likely to slowly spread, mixed by the ocean currents, until it became entrained in the Loop Current. At that point, according to the simulations, speeds picked up to about 40 miles per day; when the liquid entered the Atlantic’s Gulf Stream it could travel at speeds up to about 100 miles per day, or 3,000 miles per month. It could move north as far as about Cape Hatteras, North Carolina, with the Gulf Stream, before turning east.
The study contained a number of caveats, leading Synte to stress that the simulations were not a forecast but rather an envelope of possible scenarios. The timing and course of the oil slick would be affected by regional weather conditions and the ever-changing state of the Loop Current. The dye tracer used in the model has no actual physical resemblance to true oil. Unlike oil, the dye has the same density as the surrounding water, does not coagulate or form slicks, and is not subject to chemical breakdown by bacteria or other forces. Mitigation efforts such as physically skimming off the oil weren’t included; for this reason, the extent of dye in the model likely overestimates the degree to which oil would spread in the ocean.