The cataclysmic earthquake that struck off Sumatra in December 2004 caught seismologists by surprise (Science, 14 January 2005, p. 201). It just wasn't the sort of place they expected a long fault to break all at once. Now researchers are realizing that the same fault poses a threat all around the Bay of Bengal, from Myanmar to Bangladesh and India. The hazard remains highly uncertain, but "you could lose millions of people up there in one earthquake," says seismologist Kerry Sieh of the California Institute of Technology in Pasadena.
The unsettling realization comes after an analysis of recently published geological and geophysical data, historical accounts from the region, and modeling of tsunamis. In the 6 September issue of Nature, seismologist Phil Cummins of Geoscience Australia in Canberra argues that the Indian tectonic plate is not harmlessly sliding by the adjacent plate along a fault far inland in Myanmar, as previously assumed. Instead, the Indian plate is diving beneath the other plate well offshore of Myanmar, buried beneath the thick sediments of the Bay of Bengal. Any large quake there could generate a tsunami, Cummins notes.
Such a quake would not be without precedent. Historical reports suggest a major quake shook the Myanmar region in 1762, and Cummins's analysis backs this up. He finds that the way land rose in the south and sank in the north supports a rupture of 700 kilometers of the fault, which may have generated a magnitude-8.8 earthquake. In a model, the quake generates a large tsunami along the nearby Myanmar coast. If a similar rupture were to recur, the quake and tsunami combined would threaten Chittagong and Dhaka, Bangladesh, and Kolkata, India. Given the dense populations there, "it seems likely that the number of lives at risk may be over a million," Cummins writes.
Other researchers agree that the fault poses a risk to the Bay of Bengal; the question is the degree of hazard. In unpublished work, tsunami researcher Costas Synolakis of the University of Southern California in Los Angeles and seismologist Emile Okal of Northwestern University in Evanston, Illinois, identified a shorter rupture and therefore somewhat smaller quake for the 1762 event and thus less devastation for any future quake. Still, writes Synolakis, "the details will only affect the geographical distribution of the carnage; ... there is no more time to waste."