GEOPHYSICS

A BLUE NOTE

Seismologists find a mysteriously pure tone in the ocean

When Jacques Talandier of the French Atomic Energy Agency and Emile Okal of Northwestern University examined some loud rumblings recorded by the network of seismic stations in French Polynesia, they discovered, much to their surprise, a single frequency--in essence, a pure tone--blasting through the ocean. Was it an animal? A secret navy experiment? None of these speculations held up under scrutiny, and they were stumped. But Talandier and Okal can now rejoice with a new explanation. And perhaps they should have celebrated earlier, because opening a bottle of champagne might have helped them solve the mystery.

The "monochromatic" seismic signals that caught their attention were each composed of just one frequency--typically in the range between three and 12 cycles per second--making them purer than a note from a musical instrument, which invariably includes various overtones in addition to the fundamental frequency. (The combination of overtones present distinguishes a note played, for example, on an oboe from the same note played on a piano.) These oceangoing sound waves--called T waves--were particularly cacophonous in 1991 and the early months of 1992. Individual blasts lasted from a few seconds to several minutes. Earthquakes would have produced much more short-lived signals. Whales would have emitted higher-frequency sounds that showed seasonal changes. "This was quite different from anything we had seen anywhere else," Okal recalls.

Although similar seismic signals, called harmonic tremors, have come from the magma bodies lurking beneath some volcanoes, such resonances usually generate overtones. Perhaps, the two seismologists reasoned, it was merely a limitation of their instruments, which had been designed to filter out extraneous high-frequency noise. Talandier and Okal turned to recently declassified recordings made by the U.S. Navy, which operates arrays of underwater microphones designed to listen to higher frequencies--like those given off by submarines. But the navy's data from SOSUS (the military moniker for SOund SUrveillance System) showed that the oceangoing sounds curiously lacked high-frequency overtones.

The first clue to the solution came after the two seismologists determined the position of the source, which was within a poorly surveyed region of the South Pacific. Old charts indicated an underwater volcanic ridge in the area. And so the two researchers coaxed colleagues to visit that locale, an expedition completed last year. New probing revealed a flat-topped undersea volcano that rose to within about 130 meters of the surface. Although no volcanism was obvious at the time, the samples recovered contained fresh lava, indicating volcanic activity in the recent past.

Talandier and Okal knew that vast stretches of the seafloor are currently rife with such volcanism but that few volcanic events generate T waves. Those that do, such as the South Pacific seamount, are located at shallow depths, where the pressure is sufficiently low that bubbles can form in the water above the scorching lava. So the source of the curious T waves seemed linked in some way to undersea effervescence.

Searching for further insight, they consulted Bernard Chouet, a specialist on harmonic tremors at the U.S. Geological Survey, who urged them to consider the interesting things that can happen in a mixture of water and steam. For example, sound waves, which typically travel about 1,500 meters per second in the ocean, can go as slowly as one meter per second. "You can walk faster than that," Chouet quips. He imagined that the resonator in this case was probably a cloud of bubbles sandwiched between the top of the seamount and the surface of the ocean.

Chouet ran computer simulations to see whether such a cloud might behave as a resonant cavity--acting much the way an organ pipe does when it sounds a note. He found that sound waves would indeed shoot up and down through the cloud at some resonant frequency, reflecting back and forth between the ocean surface and the seamount. But little energy would bounce sideways, because the diffuse boundary of the cloud would not produce reflections. As a consequence, the fundamental frequency would remain steady, no matter what the lateral extent of the cloud. This bubbly body, like a musical instrument, would also generate overtones, but there would be a natural tendency for the gas bubbles to damp out the higher frequencies.

In their report in the Bulletin of the Seismological Society of America last year, Talandier and Okal presented this resonating bubble cloud under the heading "Volcanological Speculations." So they are perhaps not entirely convinced themselves. And although a resonating slab of frothy seawater seems a neat explanation, Chouet warns that "anything is possible."

--David Schneider