When Bill Gates got married on the Hawaiian island of Lana'i, he might have wondered how the coral-rich gravel outcrops shown here came to be 30 metres above sea level. If so, he would have been told that these gravels, and others even higher up, were deposited by a giant tsunami 105,000 years ago.

The tsunami hypothesis came from the chaotic nature of the gravels, and the fact that other nearby Hawaiian islands are subsiding — so that deposits formed at sea level 100,000 years ago would not be expected to be at high elevations today. But the tsunami would have had to be truly 'giant', with a run-up of hundreds of metres, compared with a modern-day Hawaiian record of only 17 metres.

Elsewhere in this issue (Nature 408, 675–681; 2000), Ken Rubin and colleagues describe a test of the tsunami hypothesis. They find that the gravels contain corals of two distinct ages (135,000 and 240,000 years old), both from times when sea level was high. From this and detailed mapping studies, they conclude that the gravels were in fact deposited not in a single catastrophic event, but by typical Hawaiian coastal processes such as those forming the cobble beach towards the left of the photo.

And what of the subsidence problem? It turns out that recent flexure models of the Pacific plate can explain concurrent subsidence of some islands and uplift on others, including Lana'i. Uplifting experiences for both the island and the computer magnate, then.