Frequency spectrogram of T wave showing relative strength of different frequencies through time

When the sea floor off the coast of Sumatra split on the morning of December 26, 2004, it took days to measure the full extent of the rupture. Recently, researchers at Columbia University's Lamont-Doherty Earth Observatory analyzed recordings of the underwater sound produced by the magnitude 9.3 earthquake. Their unique approach enabled them to track the rupture as it moved along the Sumatra-Andaman Fault, raising the possibility that scientists could one day use the method to track underwater earthquakes in near real time and opening new avenues in seismologic research.

Listen to the December 26th earthquake (mp3)

"We were able to constrain some details such as the speed and duration of the rupture more accurately than traditional seismic methods," said Maya Tolstoy, a Doherty Research Scientist and lead author of the study. "Moreover, we found the earthquake happened in two distinct phases, with faster rupture to the south and slower to the north, almost as if there were two back-to-back events." The study appears in the July/August edition of Seismological Research Letters.

The earthquake occurred along a stretch of the fault known as a subduction zone, where the India Plate is slowly being pushed beneath the Burma Plate. "The fault basically unzipped from south to north," said DelWayne Bohnenstiehl, a Doherty Associate Research Scientist and co-author of the study. "By looking at the direction the sound comes from, you can get a pretty clean look at the way it broke."

When an earthquake occurs underwater, part of its energy is released in the form of sound, known as a tertiary or T wave, which travels great distances through the ocean. T waves travel significantly slower than primary and secondary (P and S) waves, both of which are often recorded in an overlapping pattern that can obscure important parts of the seismic record. As a result, oceanic sound energy sometimes provides a more direct look at the entirety of a large underwater earthquake. "It's like the hare and the tortoise" said Tolstoy, "The tortoise is moving a lot more slowly, but it gets the right answer in the end".

The T waves for the Sumatra earthquake were captured by underwater microphones located at Diego Garcia, more than 1,700 miles from the epicenter. These microphones are part of arrays known as hydroacoustic stations that are scattered throughout the world's oceans to listen for the telltale sound of an atomic blast.

Waveform showing T wave amplitude

"We can hear icebergs cracking and magnitude four earthquakes underwater from across the ocean basin, so it's not surprising that we heard this earthquake," said Tolstoy.

What is surprising, however, is the fact that the earthquake appeared to occur in two distinct phases. The first phase encompassed the first three minutes of the eight-minute earthquake, during which the rupture proceeded north at about 1.7 miles per second (2.8 km/sec) from the epicenter. During the second phase, the rupture slowed to 1.3 miles per second (2.1 km/sec) and continued north for another five minutes until it reached a plate boundary where the fault changes from subduction to strike-slip, where the two plates push past one another in opposite directions. This suggests that had the subduction continued, this longest ever recorded earthquake might have been even longer.

The method that the researchers used also shows promise for helping officials quickly determine where relief activities are needed. In the case of the Indonesian earthquake, early seismic data indicated that only the southernmost third of the fault was involved. Later analysis revealed that about 750 miles actually ruptured, a finding that was supported by Tolstoy and Bohnenstiehl’s study.

Because hydroacoustic stations like the one at Diego Garcia operate around the clock, Tolstoy believes they may hold the potential to provide a rapid and accurate source of information on the duration and length of an underwater earthquake, information that is critical in determining where to send emergency relief in the first hours of a disaster, as well as in determining the risk of a tsunami.

Recently, the International Monitoring System of the CTBTO began making their data available on a trial basis to tsunami warning organizations recognized by UNESCO. Tolstoy hopes that eventually scientists will gain easier access to these data as well, which would help them learn more about the basic processes of the Earth. "There is an opportunity here to make a contribution to international disaster monitoring, as well as help us better understand earthquakes and tsunamis and potentially mitigate these events in the future." said Tolstoy. "It makes sense to let others listen in.