ST. LOUIS (AP) -- Scientists disagree on the chance of a major earthquake in the New Madrid Seismic Zone. But a study in the journal Nature reports that strain has been building in the New Madrid fault, the St. Louis Post-Dispatch reported Thursday.

The study's authors from the University of Memphis said pressure on the Reelfoot fault is rising, which would be required for an earthquake.

At the surface, the Reelfoot fault is a gentle swell that runs from New Madrid, Mo., to northwest Tennessee. It cracked in the huge New Madrid quakes of 1811-1812, when the west side of the fault shot up, the other side down, as much as 10 feet.

Every year, there are about 100 deep, barely detectable tremors in the New Madrid region. But so far there has been little measurable motion at the surface--at least nothing like the San Andreas fault, where tectonic plates move past each other more than an inch a year. That strain stresses the fault for future earthquakes.

The Memphis researchers placed 11 global positioning systems in the muddy soils, with some of the instruments resting on steel beam foundations more than 60 feet deep.

After four years, two stations on opposite sides of the Reelfoot fault had moved almost half an inch closer together. That motion over long periods of time could cause the fault to buckle in an earthquake.

"We're delighted that the GPS results are finally showing something consistent with the geological evidence," said geologist Michael Ellis, one of the authors. "The surface of New Madrid is moving. And it contributes a significant amount of strain."

But the results contradict the negligible motions Northwestern University geophysicist Seth Stein reported in 1999 in the journal Science. Stein used his results to argue that the New Madrid seismic hazard was overestimated.

He called for the U.S. Geological Survey to change the hazard maps that guide building codes, something Ellis says was "irresponsible." And Stein said he's not convinced by the Nature paper.

Neither are three geoscientists who have analyzed the University of Memphis data without finding significant motion.

The critics say random chance alone can explain the strain between the two Reelfoot fault stations. There needs to be a more consistent trend for all 11 stations, said Purdue University geophysicist Eric Calais.

Calais said the Memphis researchers were optimistic in their treatment of uncertainty, while he and researchers at the University of Arkansas and University of Wisconsin were more conservative.

---

Information from: St. Louis Post-Dispatch