In-Depth: Quakes Threaten Eastern U.S.

July 28, 2006 — The fortunes of a string of cities from Memphis to Quebec City may be written in a maze of Mississippi River mud and sand near New Madrid, Missouri.

Buried in the old river channels are clues to what may someday be the largest natural disaster in U.S. history, say geologists who study the deceptively serene seismic realm that is eastern North America.

Worst U.S. Disaster Ever?

It sounds like hyperbole, but after all it was New Madrid, not San Francisco, where the most powerful quakes in the history of the lower 48 states were centered.

The quakes that put New Madrid on the map — and almost wiped it off — were a cluster of landscape-scarring monsters that hammered the once sparsely populated region over a seven-week period from Dec. 16, 1811 to Feb. 7, 1812.

In the two centuries since, scientists studying the "New Madrid Seismic Zone" have usually reconstructed those quakes — which were up to 8 in magnitude — by relying on historical eyewitness accounts of collapsing chimneys, a backward-flowing Mississippi, and quicksands that swallowed whole islands.

"The New Madrid zone has the highest hazard rating east of the Rockies," says geologist Russ Wheeler of the U.S. Geological Survey in Memphis. "It's head and shoulders above anyplace else."

That’s not just because of what happened in 1811-12. There is now ample evidence of previous big quakes at New Madrid. That evidence includes contorted sediments, once-liquefied sands and, most recently, abruptly straightened river channels.

A Knotty Problem

From high above New Madrid, the Mississippi River is stamped on the land like a giant imprint of shabbily braided twine. It’s a knotty mess. But there is a method to the meandering maze, says geologist John Holbrook of the University of Texas at Arlington.

When the region is undisturbed by earthquakes, the Mississippi River wends its meandering way south. But when the Reelfoot Fault near New Madrid gives way, it can abruptly raise the land to the southwest, changing its slope.

"We think that’s what happened in 1811-12," says Holbrook. "That created a tsunami upstream."

That may explain some eyewitness accounts of the river flowing in the wrong direction. But such an abrupt change also leaves an enduring mark on the river, Holbrook says.

The new, more gradual slope causes the river to abandon its meanders and straighten just above the fault zone. After about a thousand years, the river erodes away from the bulge created by the quake and returns to a sinuous course.

That cycle, says Holbrook, ought to be detectable in river sediments. To find it, he and colleagues have spent years hand-drilling cores of sediment in the New Madrid area to reconstruct past river channels.

In the July issue of Tectonophysics they reported two cycles of straightening apparently caused by ancient, massive quakes.

"At about 2200 B.C. the river straightens out," says Holbrook. The river gradually returned to a meandering course over the next thousand years. "Then at about 900 A.D., the river straightens again and starts all over," he says.

The 900 A.D. event had already been noticed by paleoseismologists looking in the Mississippi Valley for evidence of the sudden shifts of stable sand into quicksand that can occur when it is shaken violently.

Paleoseismologist Martitia Tuttle had seen hints of the earlier event — which she had dated to 2350 A.D. — but not enough to be sure.

"That’s probably the same event," Tuttle says.

The difference is that Holbrook’s work can connect the paleoseismological clues directly to the Reelfoot Fault, the same fault that caused the 1812 disaster.

"There are liquefaction features in the area but we didn’t feel we had enough" to firmly say there was another distinct event or to connect them to the Reelfoot Fault, Tuttle says.

  Quicksand & Cracks

Holbrook’s new technique is getting mixed, but hopeful, reviews by other geologists who specialize in the New Madrid area.

"This is novel stuff with great potential," Wheeler says. "Like anything novel, it needs to be tested."

"The paleoseismologists haven’t been able to see beyond (1450 A.D.), other than hints," he adds. "What John (Holbrook) may have come up with is a way to extend the record. I sure hope he’s right."

And what if he is? What good is identifying yet another earthquake, one that occurred in 2250 B.C.?

Holbrook argues that there may be 1,000-year gaps between clusters of big events at New Madrid. The Reelfoot Fault, he says, could switch on and off over long periods.

A similar pattern has been seen on major faults in more seismically active areas at the edge of tectonic plates, like California, says Holbrook.

"When you get away from plate boundaries, there’s long been suspicion that maybe there was temporal clustering too," says Holbrook, "but over longer periods."

That could be good news for the Reelfoot Fault, which may be centuries away from another set of whoppers. But does that mean other faults, which have been quiet in the relatively short history of European occupation of North American, could suddenly wake up?

"We don’t know," says Holbrook. But it’s a disturbing implication of his work, he says.

Seismic Shutes & Ladders

The danger of big quakes in most places east of the Rockies is twofold: They affect larger areas, and because they are infrequent, older buildings that have never experienced even moderate shaking are likely to collapse if they do.

Quakes east of the Rockies affect larger areas because of the hard, crystalline rock that underlies much of the land there. This "basement" rock is essentially transparent to seismic waves.

"The waves shoot straight through," says Memphis-based USGS paleoseismologist Buddy Schweig. And when the waves do encounter soft, river or bay sediments – on which most cities are built, he points out — they are amplified tremendously.

This is in stark contrast to the western U.S., where rocks are hotter, more faulted and jumbled with mountains — qualities that slow and disperse seismic energy as it spreads from the epicenter.

The effect, says Schweig, is that eastern quakes can throttle 20 times the area of comparable California quakes.

This also applies to New England and southeastern Canada, where rare strong earthquakes are felt at extraordinary distances. In 1755, for instance, the Cape Ann Earthquake caused damage from Nova Scotia to South Carolina, says Boston College seismologist John Ebel.

"It was very typical of moderate ground shaking," said Ebel. Yet there were so many fallen bricks in downtown Boston that the streets were impassable to horses. The same quake today, of a magnitude of about 5.5, would probably cause billions of dollars in damage to the greater Boston area, says Ebel.

And there have been other earthquakes in New England and southeastern Canada, says Ebel. There was a big one in 1638, another in 1663 and then again in 1727. Even today there are tremors, he says.

"We see really small earthquakes throughout the region, including some damaging earthquakes," says Ebel, who was trained as a seismologist in far more active California. It’s enough, he says, to give him an inkling of the overall pattern.

"What I discovered is that everything I learned in California I was seeing here," Ebel says. "The only difference is what you see in one year in California, you see in 100 years in New England."

The good news is that cities from Boston to Memphis are taking the threat seriously. The work of people like Schweig and Tuttle and soon, perhaps, Holbrook is allowing the creation of better hazard maps. These, in turn, are leading to sturdier bridges and tunnels and tougher new building codes.

"I’m sure the Earth is trying to tell us something," says Wheeler, referring to the history of big earthquakes in otherwise quiescent places. And via those Mississippi muds, we may be starting to get the message.