The generation of submarine fiber-optic cables that revolutionized trans-oceanic telecommunications a decade ago is being retired prematurely. Dramatic advances in optical technology and a glut of fiber capacity make these cables uneconomical for telecommunications. The fibers will not go dark entirely, though: a nonprofit science group says the obsolete cables can be a boon for undersea seismology and oceanographic research. The cables, laid from 1988 to 1993, were designed to operate for 25 years; all but one are in working order.

The undersea optical fibers dramatically increased the capacity for telephone traffic across the Atlantic and Pacific when they began operation. The first of them carried 280 megabits per second on each of two fiber pairs, the equivalent of 35,000 telephone circuits. That was an impressive total at the time-but newer cables have hundreds or thousands of times more capacity. In March, Tyco Telecommunications reported that each of the eight fiber pairs in a new 9,000-kilometer cable between Oregon and Japan could carry 960 gigabits per second, giving it a total capacity more than 10,000 times that of the first fiber cables.

“The owners are basically rationalizing the stocks of cable,” explains David Robinson of the sub-sea business group at BT, the former British Telecom. With plenty of extra capacity on new cables, BT and other companies that shared ownership of the first transatlantic fiber cable didn’t bother to repair the cable when it failed in late 2001. As they had earlier retired the seven copper cables that preceded the optical one across the Atlantic, last year the companies quietly shut down the fiber cable, known as TAT-8. BT and its partners will soon retire three other early fiber cables, TAT-9, -10, and -11. But the cables that blazed a new path for telecommunications could find a new life in scientific research. 

The Washington-based Incorporated Research Institutions for Seismology (IRIS), a consortium of universities that collect seismic data to study the Earth’s interior, wants to adapt the working cables to serve sea-floor research stations. AT&T, a partner with BT in the TAT cables, gave IRIS a retired Pacific copper coaxial submarine cable in 1998, and is willing to give the group its share of the old fiber-optic cables. The fiber cables can transmit hundreds of megabits of data from seismic stations and other automated sea-floor observatories, says Rhett Butler of IRIS, whose day job is managing the National Science Foundation’s global seismic network. The stations can also tap the kilowatts of electric power that the cables carry for submerged electronics. “Our first goal is to acquire these for the scientific community,” says Butler. With two-thirds of the planet under water, he adds, “we’re going to have to have sea-floor observatories.”

Transfer of the fiber cables has hit a snag in European regulations that require removing old cables from national waters inside the 20-kilometer limit.  BT and other operators are proceeding with plans to pull up the cables. Robinson says he has not received a formal proposal from IRIS. The company will consider requests, but wants to be sure the new owners assume liability for removing the old cables.

European telephone companies have been removing their ends of transatlantic cables since they retired the first such transmission line, TAT-1, in 1978. Installed in 1956, TAT-1 sent electrical signals through coaxial cable, with vacuum-tube amplifiers spaced along the cable to amplify its 36 telephone circuits. Engineers improved undersea coaxial cables for two decades, replacing the vacuum tubes with transistors, but eventually reached a limit of 4,000 telephone circuits on TAT-6 and -7, installed in 1976 and 1983. Communication satellites looked set to drive submarine cables out of business until fiber optics came on the scene. Teams at Bell Telephone Laboratories (then part of AT&T) and British Telecom Research Laboratories made a risky bet on a new kind of fibers, in which the light-carrying core measured a mere 9 micrometers in diameter-six times smaller than the cores used in early terrestrial fiber systems. These new “single-mode” fibers offered higher bandwidth, but aligning the light-carrying strands with one another required extreme care. Submarine cable developers overcame that challenge so well that by the mid-1980s, single-mode fibers became the standard for long-distance transmission on land.

The downfall of the first generation of submarine fiber-optic cables was their need for repeaters-devices that boosted the signal strength periodically to enable the information-carrying light waves to span the whole ocean. Early fiber repeaters had to convert faint optical signals to electronic form so they could be amplified, then convert the electrical signals back into light. In the late 1980s, a new kind of optical fiber was developed that could amplify weak optical signals switched on and off 10 billion times a second or more. Better yet, they can simultaneously amplify signals at several different wavelengths (a technique known as wavelength division multiplexing). Those breakthroughs allowed cable manufacturers to build global networks with transmission capacities that dwarfed the old TAT-8, -9, -10, and -11 cables, and made them uneconomical for carrying telecommunications traffic. But if scientists can resolve procedural snarls on the European end, the old cables will have a new life helping them explore the ocean depths.