In April, when the Deepwater Horizon petroleum drilling rig exploded and oil began gushing from a drill hole almost a mile deep in the Gulf of Mexico, scientists and engineers scrambled to figure out where the oil was going. How much was staying in the depths? Where would it be in a day or a week or a month?
The technical challenges of tracking undersea plumes of oil are huge. Instruments called current profilers can detect currents flowing hundreds of meters below the surface, but until recently they had to be lowered from a ship to their target depth, left in place long enough to get a reading, and then hauled back up so the data could be retrieved—a laborious process not well-suited to an urgent situation.
Now, thanks to a chance encounter and some imaginative tinkering, scientists from Woods Hole Oceanographic Institution (WHOI) will be able to take real-time readings of currents and oil plumes deep in the Gulf. The key to the technological breakthrough is the humble sea cable.
For decades, sea cables used by oceanographers to lower devices into the ocean also served as modest pathways to transmit simple electronic signals between ships and instruments dangling miles below. But in recent months, the ordinary sea cable has become an unlikely star in the hands of WHOI engineer Marshall Swartz. Using mostly off-the-shelf computer hardware, he has built a device that turns sea cables into undersea data superhighways. (See profile of Marshall Swartz.)
Rather than a trickle of data, the sturdy, reliable, and relatively inexpensive cables can now transmit torrents of information—photos, video, sonar, chemical, and other types of data, all in real time. Scientists are already calling Swartz’s development a “game changer” that dramatically improves their ability to work at sea.
Swartz’s device, known as the SDSL Data-Link (Synchronous Digital Subscriber Link) provides an enormous advantage, scientists say—the chance to keep their instruments working in the water as they monitor incoming data. Before, many scientists were forced into a time-consuming pattern of hauling their instruments back aboard ship, downloading data to see what they had, then lowering their gear again.
“It blows open a realm of possibilities for more efficient field work,” said geologist Dan Fornari, who has pioneered and built several deep-sea camera systems at WHOI’s Deep Submergence Laboratory.
Last fall, Fornari helped deploy a prototype of the Data-Link system in the Guaymas Basin, an undersea valley about 2,000 meters (1.24 miles) beneath the surface of the Gulf of California. WHOI scientists Dan Lizarralde and Adam Soule had originally set out to use underwater cameras attached to a high-speed, high-bandwidth fiber-optic cable. Their goal was to take photos to make undersea maps and locate marine life in an area where they hypothesized warm, carbon-rich fluids were seeping through the seafloor and supporting biological oases on the mostly barren mud.
But when a system that operated the fiber-optic cable needed repair, they agreed to try Swartz’s Data-Link sea cable system.
"It worked perfectly the very first time," said Lizarralde.
“Without it, we were stuck,” Soule said. “We needed that camera to see things in real time. It was essential to what we needed to do.”
Using the SDSL Data-Link, the camera on the end of a sea cable sent about 15,000 images up the cable in a 48-hour period. Among them were shots of huge white mats of bacteria, red anemones, and Frisbee-sized starfish. As Soule watched the images flow in, he used the data to redirect the ship.
“It was like we used a heat-seeking missile and just zoned right in to our seafloor target,” he said. “It was a rare, rare case where we went to sea to find out something, and we found exactly what we were looking for.”
Fornari was so impressed that he commissioned Swartz to build two more systems this year to use on research cruises to explore hydrothermal vents in the Pacific.
"It makes no sense not to have one of these things with you, if you have any notion that you might want to look at the seafloor," said Lizarralde.
When sea cables were developed three decades ago, they were designed primarily to lower and lift heavy instruments. Inside the cables’ flexible metal skin, engineers nestled copper wires that could send electronic signals down the cable to trigger simple tasks, such as turning on a camera or closing a water sample bottle. But the system had limited data and electrical capacity.
“We just never thought twice about using sea cables for anything requiring high bandwidth,” Swartz said.
As Internet technology developed, oceanographers leaped at the chance to use fiber-optic cables, which can transmit huge data files—fast. But with that speed came higher cost and considerable inconvenience. On ships, fiber-optic cables are heavy and require special winches. Trained engineers must install and maintain the systems, and their operation takes up an enormous amount of deck space, something few ships in the American fleet of oceanographic research vessels have.
Currently, just four ships in the UNOLS fleet of 26 large oceanographic research vessels have fiber-optic cables on board. Swartz estimates that installing fiber-optic cables on the rest of the ships would cost about $2 million per vessel.
But 17 of those same ships have sea cables on board, and he can build his Data-Link for about $15,000 per cable. His goal was to make it easy for ship technicians to set up and use the Data-Link on a cable with no new required winches, wires, or other supporting equipment. Also, it doesn’t interrupt any of the cable’s originally intended uses.
“I deemed it necessary that in order for this to be a successful product to our science pool, we had to be able to use it without modifying existing equipment,” he said. “I found I could do that with most things that came right off the shelf.”
Swartz built the prototype Data-Link in his lab for about $5,000. The key was building in an Ethernet system, which directs the flow of data over a shared cable, not unlike the way a policeman can prevent collisions and keep more traffic moving at intersections. Only a decade or two ago, most people connected to the Internet using an Ethernet, which handled increasing amounts of data—from e-mails to shopping orders and images on Web pages—through a network of copper-lined cables maintained by telephone companies.
Swartz said he simply applied existing technology to sea cables. On the seafloor end of the cable is a watertight, pressure-resistant, orange can with switches, converters, and adapters that enable scientists to plug in their various instruments. Data sent up the cable are made available on Ethernet by a converter in the ship’s lab.
“There’s no rocket science here,” Swartz said. The most expensive and technically challenging aspect of the project was developing a power supply box, a job that went to WHOI electronics engineer Steve Liberatore.
Swartz readily admits that fiber-optic cables are still the superior system for demanding applications such as high-speed video. If fiber-optic systems are bullet trains, his is more like a bus; but buses are more common and less expensive than bullet trains, and they still get you where you need to go for many types of work.
Swartz realized the sea cable’s potential as an Ethernet carrier over Thanksgiving weekend in 2008, when he stopped by the research vessel Knorr, docked in Woods Hole, to make a few fixes on an instrument. As he walked through the ship, he spotted University of South Florida scientist Drew Remsen, who looked anxious and miserable.
“He was in a panic because he was leaving the next day for a cruise, and he had no one to help him,” Swartz said. Remsen explained that he was trying to attach a video camera to a short sea cable so that he could receive real-time images of the ocean below the ship. A consultant in Florida had helped him bring the parts together for the cable; Remsen just needed help fitting the pieces together.
Initially skeptical, Swartz looked closer at what Remsen proposed and saw that the sea cables seemed to have all the right stuff to make a high-speed Ethernet connection possible.
Sea cable is made up of three twisted copper wires wrapped inside steel armor. But oceanographers typically use just one of the three wires as a pathway for transmitting electrical power and signals to their instruments. Swartz said that the other two serve simply as a back-up system if the first wire fails, which rarely happens.
A lightbulb went on in Swartz’s head.
“So my thought was, “Gee, if this works, we can just attach the Ethernet connection to the extra wires. I said to Remsen, ‘We’ve never seen anything like that, but let me try something.’ I plugged it in. It worked. I thought, ‘Oh my gosh, this shouldn’t work.’ It countered everything we had ever experienced. It was just like something had landed from outer space. No one expected it, and it was groundbreaking because it allowed information to absolutely fly along standard sea cables.”
Swartz stayed on the ship experimenting with the cable until four in the morning. The next day, still bubbling with excitement, he explained to colleagues what he had seen. What if he could build a similar system for use on longer sea cables? Remsen’s cable was less than 1 kilometer long. WHOI sea cables were typically many miles longer and ran the risk of losing a signal, because signals typically degrade over longer distances.
“We first checked it out on 7.4-kilometer (4.6-mile) sea cable outside my lab,” Swartz said. “We plugged it in, and bingo, it worked. I thought, ‘Man, what else could we do?’ ”
He started a battery of tests. In a test of video-gathering capabilities, he used the Ethernet connection on the cable to watch CNN on his computer. He put his feet on his desk and called Fornari, the operator of deep-sea camera systems, to come over to his lab.
“He shows up and says, ‘You asked me to watch you watching TV on your laptop?’ and I said, ‘It’s running through the sea cable.” His jaw just dropped. He said, ‘Oh my god, we could use that with our cameras.’ So immediately he was on board.”
Six months of work followed to develop and test the Data-Link. In August 2009, WHOI scientist Rich Camilli was the first user of the system during a cruise off Santa Barbara, Calif. He was looking for cold seeps where oil and other naturally occurring hydrocarbons are leaking out of the seafloor.
Following Swartz’s instructions, Camilli hooked up a sea cable to a watertight power box and about five pieces of off-the-shelf computer communications devices—the same electronics used to establish Ethernet connections on home or business computer networks. Then Camilli installed an undersea mass spectrometer called TETHYS, an instrument that measures minute quantities of chemicals in seawater, on the cable and lowered it into the ocean from the deck of the research vessel Atlantis.
“Bang, it worked right out of the box,” Swartz said.
Ten minutes later, when Camilli returned to a lab on the ship, he was stunned to see measurements streaming in of methane in the seawater below. Based on the volume of data he was seeing pop up in real time on his computer screen, he could redirect the ship’s course to home in quickly on his scientific targets.
“It was a game changer for me,” Camilli said.
The new system could also be a game changer for scientists trying to track plumes of oil from the Deepwater Horizon disaster. This month Camilli, Swartz, and the new Data-Link are heading to the Gulf of Mexico on the R/V Endeavor, a National Science Foundation ship operated by the University of Rhode Island. They will use the Ethernet system with two key instruments: a Lowered Acoustic Doppler Current Profiler, which will record information on strength and direction of currents; and Camilli’s undersea mass spectrometer, which will detect chemicals from the oil spill.
“This system is the only way to get real-time data on the plumes,” said Swartz. “Our intent is to understand the behavior of the water mass around the plumes”—and gain a clearer picture of a very messy situation.