ANIMALS and plants are not the only things that form fossils. Tsunamis—the huge waves created by some submarine earthquakes—do so, too. A tsunami generated in January 1700, off the Pacific coast of North America, for example, has left abundant traces in local rocks, as well as in the art of Japan (it was the inspiration for Katsushika Hokusai’s woodcut, “The Great Wave of Kanagawa”). That should give pause to coastal dwellers in what is now Oregon, Washington state and British Columbia. They might reasonably wonder when the next big wave will arrive—as might residents of other earthquake-prone coastlines around the world.
Harvey Kelsey, of Humboldt State University, in California, has done more than wonder. He and his team have been looking at the Indian Ocean coast of Aceh province, in northern Sumatra. This was the origin, in December 2004, of a powerful submarine earthquake and subsequent tsunami that killed around 230,000 people. The quake in question, as is commonly the case for quakes of this magnitude, was caused by one of the Earth’s crustal plates sliding under another. Plate movement at such boundaries tends to be episodic, with strain that has built up over the years being released suddenly as a tremor, when it becomes too much for the rocks to sustain. The rate at which the quake-causing strain accumulates, however, is constant. This means earthquakes of this type can be expected at reasonably regular intervals.
As they report in Geology, Dr Kelsey and his team looked for evidence of past tsunami-causing earthquakes in Aceh, going back about 8,000 years. To do so they collected cores of sediment from the local sea bed—a dozen in all—from two places close to the modern shore line. As might be expected, given their provenance, a fair amount of each of these cores was composed of sand. Interspersed with the sand, though, were three layers of soil rich in mangrove pollen and in fossils of tiny animals (called foraminifera, or forams) of species that are commonly found in mangrove forests today. Each of the three mangrove layers was abruptly terminated by sand that had clearly been laid down in turbid conditions, rather than settling gently. This material was also loaded with forams, but these were species normally found only in the open ocean.
The implication is clear. Mangrove forests grew and flourished, and were then swept away by tsunamis. Radiocarbon dating of the top of the mangrove layers shows these tsunamis happened 7,000, 5,800 and 3,800 years ago. Moreover, close scrutiny of the cores revealed four other turbid layers that had not disrupted forests, because none was growing in the area at the time. From 8,000 years ago to 3,800 years ago, then, there seem to have been seven tsunami-generating earthquakes.
This older sediment was able to build up in an orderly, stratified fashion because rising sea levels after the last ice age meant it was never too close to the waves at the surface. By 3,800 years ago, though, sea-level rise was over, so in younger sediment wave action has mixed up the strata. Even so, Dr Kelsey was able to estimate what proportion of this sedimentary mix had arrived in tsunamis by looking at the proportion of forams in it that were oceanic. Knowing this, the thickness of the sediment in question, and the average thickness of a tsunami fossil, he estimated that between four and six tsunamis have struck the area in the past 3,800 years. All told therefore, the geology of the region suggests that between 11 and 13 tsunami-causing earthquakes have hit it in the past 8,000 years. This works out to one every 620 to 730 years.
For Aceh, to know this is to shut the stable door after the horse has bolted. There is no reason, though, why Dr Kelsey’s method could not be put to useful work in places that have not suffered recent tsunamis, but might do so in the future.