Tectonic plates ceaselessly carry continents around the world, but never faster than about 10 centimeters a year. Now comes evidence of a shift 10 times more rapid: 84 million years ago, two researchers claim in today's Science , the whole Earth rolled like a ball, turning 15° to 20° in just a few million years. If the twist took place, it would have jerked the whole globe into new climate zones.
Earth's poles are a wobbly crew. The magnetic poles, for example, ramble with the vagaries of the churning core that produces the magnetic field. But despite all this wandering, experts have disagreed for decades about whether the geographic poles ever rapidly shifted their position. The physics suggests that this motion, called true polar wander, could happen. That's because a spinning Earth is most stable when its most massive parts, such as an ice sheet on its surface or a lump of particularly dense rock within it, are at the equator. If the mass forms or moves elsewhere, Earth will roll to bring it to the equator.
Any such motion should be recorded in ancient lava. Paleomagnetician William Sager of Texas A&M University in College Station and geochronologist Anthony Koppers of Scripps Institution of Oceanography in La Jolla, California, have compiled 27 ancient pole locations dated to between 120 million and 39 million years ago by radiometric argon-argon dating. They assumed that each of the Pacific Ocean seamounts--mountain-sized piles of lava--had locked in a single magnetic field whose orientation points to the pole's location at the time the lava solidified.
Sager and Koppers found an odd situation about 84 million years ago. Seamounts from that period, give or take 2 million years, yield two pole locations 16° to 20° apart. By their analysis, the area of the Atlantic Ocean would have moved as fast as 110 centimeters per year southward. That would be enough to rotate Washington, D.C., into the tropical latitudes of the Caribbean today. To paleoclimatologists, such a rapid whole-Earth tumble would trigger climate shifts that seem to have come out of nowhere.
Not everyone finds the evidence convincing. "I think they underestimated the effect of data selection," says paleomagnetician John Tarduno of the University of Rochester, New York. He notes that the 84-million-year event is defined by just four pole locations, and shows up only in the Pacific while true polar wander should appear worldwide.