WITHIN MINUTES OF ELIZABETH HOLLERAN AND Atkins’ arrival at the earthquake center, Guy Thompson had started an analysis of the seismic data they’d brought back from the epicenter near Blytheville. Two hours later, Thompson’s team of USGS computer specialists had completed some preliminary modeling on deformation—how much the earth’s crust had been pushed up or down by the tremendous quakes and their aftershocks.
The GPS data Atkins and Holleran had retrieved from the building on the city’s riverfront was combined with several other GPS sites in the Mississippi Valley. Two of them—one near Louisville, the other just north of Jackson, Mississippi—were able to transmit their raw data by radio signal to receiver towers that had survived the quake. This information, along with radar interferometry readings taken by the SIDUSS satellite system, had been relayed to the USGS Earthquake Information Center in Boulder, then back to Thompson’s computer through a satellite hookup.
A little before midnight, the weary seismologists gathered in the library annex building. There were ten people, including Atkins and Holleran. A gas-driven emergency generator provided the electricity. Paul Weston, chairman of the Seismic Safety Commission, ran the session. He was accompanied by his two assistants, Stan Marshal and Mark Wren. Whenever he saw him, Atkins was struck by Marshal’s size. The guy looked like a professional boxer who’d hit fifty and spread out. Not a geologist. He had a blocky, heavy build that stretched his jacket at the seams. He never smiled. Never.
Wren was younger, easygoing. He was always carrying a laptop.
Atkins wasn’t pleased to see Weston, who was always fidgeting with his clothes, pulling lint from his trousers, or trying to straighten a crease. His hands were always moving, always fluttering around his clothing. It made Atkins nervous just to watch him. He didn’t trust the man and wanted to talk to him about the cracks at Kentucky Dam. Based on what he’d seen inside the wall of the dam with Holleran, Weston had been deliberately misleading about their size and severity during that public meeting in Mayfield. Atkins was eager to pursue the matter, but knew this wasn’t the time or place.
Guy Thompson was the first to speak. By then he’d been working more than fifty hours with very little sleep. He’d changed into a fresh Western shirt and jeans. This time he wore a buckskin shirt with tassels that hung from the sleeves. His face was haggard. He hadn’t shaved and was rapidly growing a thick black beard that matched his long, jet-black hair. In the few hours since Atkins had given him the seismic data he and Holleran had collected, Thompson’s whole demeanor had changed. His engaging smile and forceful voice had vanished. Usually the picture of buoyant self-confidence, he was unusually subdued.
As soon as Thompson cleared his throat and began to talk, Atkins realized what was wrong with him: the man was scared.
“We’ll go over the GPS and interferometry data first,” Thompson said. He started with a matter-of-fact description of how the data were transmitted.
“We began downloading at 4:00 P.M. local time. The GPS data was from the Block II constellation. The interferometry images came from SIDUSS.” The Synthetic Aperture Radar dual satellite system provided real-time images from two satellites operated in tandem in the same orbit.
“The data were transmitted on two L-band frequencies. The Y-code was in effect for antispoofing control,” Thompson said, explaining that antispoofing guarded against any fake transmissions of satellite data. The use was justified, he said, based on the extreme importance of the information.
Weston interrupted. “Let’s get to the summary, please. What kind of deformation do we have?”
Atkins sensed that Thompson was proceeding slowly for a reason.
“The GPS Master Control Station at Falcon Air Force Base in Colorado affirms the transmission,” Thompson said, ignoring Weston. “We’re concerned with two orbital planes, both focused on North America, specifically on the Mississippi River Valley.” He paused to check some notes. It looked to Atkins as though he was holding on to the desk for dear life.
“You asked about deformation. The surface deformation is phenomenal. Based on GPS data of six months ago, the satellite measurements show the ground was pushed up as much as seven feet across wide areas in the fault zone.”
Atkins was dumbfounded. There were murmurs of disbelief, gasps. That kind of uplift was unheard of. During the Armenian quake in 1988, uplifts of just over two feet had occurred across a 200-square-mile area and that was considered severe.
“With a deformation like that, you’ve got to wonder how much energy is still locked in the fault system,” Holleran said.
That had always been the key question for Atkins. They were finally getting at the answer. The deformation was staggering. For the first time since the 8.4 earthquake, he realized that Holleran had been totally right in wondering whether they were experiencing a pattern of foreshocks, not aftershocks. He looked at her and caught her glance, a nervous half smile. They had to consider the possibility they were well into a cycle leading up to yet another powerful earthquake.
“It’s quite possible there’s very little energy left in the ground,” Weston said, jotting notes on a piece of graph paper. “The elasticity in the rock may actually have decreased.”
“And deformation isn’t a foolproof indicator that we’ve got huge amounts of strain energy building up,” said Stan Marshal.
“But how do you explain the phenomenal number of aftershocks we’re having?” Holleran said. Like everyone else in the room, she knew the real test would come with the next set of GPS and interferometry readings. If they showed any additional rise in topography, it would mean seismic energy was still loading up in the fault.
“After a great earthquake, that’s entirely routine,” Weston said. “The aftershocks lasted for weeks after the Northridge quake.”
“That was a magnitude 6.7 event. This one was 8.4,” Holleran said. “I don’t think you can compare the two. It’s like comparing a one-story building with a 350-story skyscraper.”
“We’re getting ahead of ourselves,” Thompson said. “The deformation zone covers roughly 340,000 square miles. It runs east on a line extending from Blytheville, Arkansas, into portions of Kentucky, Tennessee, extreme southern Missouri, Illinois, and Ohio.” He darkened the lights and turned on a laptop, which projected a map of the Mississippi Valley on a movie screen.
“We’ve done some two-dimensional modeling of the deformed sections of the crust,” he said.
The images were overlaid on the map. The first in the series showed the epicenter at Blytheville, which appeared as a pronounced bulge.
“You can see the asymmetric dome-shaped uplift there,” said Thompson. The sharply defined upthrust of the earth spread out around the epicenter for a hundred-mile radius.
Atkins suspected he had another bombshell to announce.
“We’ve got a lot more to consider here,” Thompson went on, still taking it slowly and methodically. “The seismic data John and Doctor Holleran brought back from Blytheville corresponds with other reporting stations. We know the quake opened another fault in the New Madrid Seismic Zone. We’re still analyzing it, but this branch appears to run from just north of Caruthersville, Missouri, into northeast Tennessee and then up into Kentucky. It covers roughly 170 miles.”
If Thompson’s data held up, it meant the New Madrid Seismic Zone had effectively doubled in size. No one spoke. Everyone was too shocked to respond.
Thompson showed the new fault on a map projected on the wall. It appeared to intersect with another major fault segment, the hatchet-shaped top of the NMSZ.
Thompson’s voice had become calmer, almost detached. “The New Madrid Seismic Zone has increased from a series of connected faults roughly 125 miles long to a more complex system that extends for over 400 miles.”
Holleran looked at Atkins. Neither of them had experience with anything like this.
Thompson displayed another image, a map of southwestern Kentucky dappled with a series of dots. Each represented the location of major aftershocks along what Thompson had begun to call the “Caruthersville Fault.” Some were in the magnitude 6 range.
The aftershocks had been recorded by USGS seismic stations at Golden, Colorado; Reston, Virginia; and other locations. Seismographs as far away as Tokyo had also monitored them.
Thompson said, “We’ve been averaging about six hundred aftershocks a day. Most can’t be detected physically. The bigger ones have been bunched along the Caruthersville Fault.”
Holleran knew that Northridge, California, had experienced more than a thousand aftershocks a day for about a month, but they weren’t as strong as these, nowhere near it.
“You’ll note the bigger dots represent the magnitude 6 quakes,” Thompson said.
Holleran counted at least six of them.
Thompson wasn’t finished with his disturbing sound-and-light show. He punched some keys on the laptop and projected a series of thirty color-enhanced images. The sequence showed how the slippage had radiated out from the quake’s epicenter near Blytheville. Each image represented a second of time.
“You’ll see that the rupture didn’t occur instantaneously or proceed uniformly across the entire fault plane,” Thompson said. “It traveled in a northeasterly direction at about four kilometers a second. Some parts of the plane showed major slippage. Other parts little or none.”
The areas where the most slippage had occurred were called “asperities.” Seismologists had long paid close attention to asperities as the source of energy pulses that reached the surface at different times and places as the earthquake progressed. More specifically, they were areas in a fault that contained the most slippage. The direction and manner in which a big quake ruptured across the fault greatly affected the intensity of the ground motion. The movement was never uniform or instantaneous.
The location and timing of the aftershocks showed, better than any other indicator, the true scope and breadth of the fault.
As he stared at these clusters, Atkins’ uneasiness increased. It was the second fault connected to the New Madrid system to be discovered in less than four days. The first one had revealed itself after the magnitude 7.1 event and extended south of Memphis.
And now this.
Thompson’s computerized images reemphasized for Atkins one of the major facts that distinguished the New Madrid zone from all others he’d studied: its incredible complexity.
“We’re talking about a multiple event,” Atkins said. “If you try to break it down, you’ve got a seven-foot deformation. My God, I don’t think even Chile had anything like that.” The 1960 quake, the largest one in modern times, registered a magnitude 8.6 on the Richter. “Then you’ve got the dip slip and strike slip subevents on two different fault segments, both previously unknown. And both of them connected to the major fault system. I’ve never encountered that before.”
Thompson displayed an image that illustrated the kind of faults Atkins was talking about. Strike slip faults were primarily horizontal in their shearing movement. Dip slip faults moved down or upward. One of the images showed the distinctive horst and graben effect produced by the faulting process. A graben was a fault block that subsided or dropped down. A horst was one driven upward.
One detail continued to worry Atkins the most: the possibility that the big quakes were forming new faults deep in the earth or bringing old ones back to life.
“It’s not just the complexity and enigma of the events that scares me,” he said. “It’s the way these new faults have opened up. If there’s enough seismic energy left in the ground and one of them goes off, there’s no telling how far the damage will spread.” He reminded everyone of the duration of the 8.4 mainshock. “Over three minutes… I still have to take a deep swallow whenever I think about exactly how long that was.”
Thompson showed another slide, a map of the two new faults that had appeared during the magnitude 7.1 and 8.4 earthquakes and several other major faults that extended across portions of the Mississippi Valley.
“Notice how the 8.4 fault line extends up to the Shawneetown-Rough Creek System,” he said. The image showed how that fault, in turn, abutted two others—the Cottage Grove system, which cut across the bottom of southern Illinois, and the Wabash Valley Fault, running up along the Indiana-Illinois border. Branching off Cottage Grove was the long arm of the Ste. Genevieve Fault System, which started in eastern Missouri and followed the course of the Mississippi River down roughly to its juncture with the Ohio River.
“There it is, ladies and gentlemen,” Thompson said, stepping back to look at the screen. “I hope that worries you half as much as it does me.”
Stan Marshal immediately objected. “We don’t have any physical evidence that these faults are connected,” he said. “And we’re in no position to suggest that an earthquake on one would trigger one on another. That’s way too speculative.”
“The real issue is how much crustal shear strain is left in the ground,” said Mark Wren. “We don’t have enough data yet to run those kinds of projections.”
Wren seldom spoke at these sessions. He seemed like a competent geologist but was overly deferential to Weston, Atkins thought. He had to admit Wren was right. It all came back to getting more satellite readings to measure any new deformation.
Still, the data made him nervous. The new fault line was incredibly active.
Walt Jacobs had been largely silent up to then. It looked as if he hadn’t changed his denim shirt in days. He was withdrawn, moody, which wasn’t like him. Atkins was starting to worry about him. He knew Jacobs was sick with fear about his wife and daughter. He’d heard nothing from them since the quake, and he was still waiting for word from the two graduate students he’d sent to look for them.
“We need to consider the possibility we may be having a repetition of the 1811-1812 events,” Jacobs said. He spoke slowly, deliberately, and immediately drew a sharp response from Weston.
“We don’t have the data to support that even as a serious hypothesis,” Weston said with a flash of anger. He was supported immediately by several other seismologists. They knew he was right, and all were uncomfortable with raising the specter of the big quakes from the last century. There were shouts that Jacobs was out of line.
Continuing as though unaware of the interruption, Jacobs said, “It all happened before—the sudden emergence of new faults, lingering, violent aftershocks, a huge deformation over a vast region. The complicated pattern of ruptures, main shocks, and after shocks. The same thing that’s happening right now.”
“The mainshock must have created the waterfall on the Mississippi,” Holleran said.
“It’s still there,” Jacobs said. “We just downloaded some aerial footage from one of the television networks. The scarp looks thirty feet high.”
“It was closer to forty the night before last,” Atkins said.
“It could be eroding.” Jacobs said. “It happened in the 1811-1812 sequence.” He turned on the lights and pulled down a wall map that showed the Mississippi River, twisting like a snake with a dark line drawn across it just below New Madrid, Missouri.
“That’s where the waterfall was reported after the 1811-1812 earthquakes. You’ll notice Caruthersville, where John and Elizabeth crossed the river. It’s about fifty miles downstream, the place where the new fault breaks off into Tennessee. The waterfall in 1812 was created by a thrust fault. It was probably the greatest mid-plate thrust earthquake we’ve ever had—until the one yesterday.”
Holleran had never been so tired in her life. And yet she found herself wide awake, totally focused on the discussion. “What if that magnitude 7.1 event we had a few days ago wasn’t the first earthquake in the sequence?” she asked. “What if it was just a foreshock?”
Atkins remembered they’d argued over that very point when they were huddled in the Explorer near Blytheville. He wasn’t about to argue with her this time. The GPS data had convinced him that what she was suggesting was a real possibility.
Jacobs had already considered it. “I don’t think that’s likely,” he said, becoming more animated. “A magnitude 7.1 earthquake is one hell of a foreshock.”
Unwilling to let go of the idea. Holleran said, “One of the toughest issues we deal with is trying to figure out if an earthquake is a foreshock or an aftershock. We don’t really know for sure until you get a major earthquake. The magnitude 7.1 event, it seems to me, could easily have been a foreshock to the big quake we had yesterday. But say I’m wrong and you’re right, Walt. It could mean we’re in for one more big one instead of two. Either way, it’s a disaster.”
“I don’t want any more talk about a triple,” Weston said angrily. “You’re both bordering on irresponsibility.”
Again, there were loud murmurs of support. Most of the seismologists in the room, like Weston, were convinced that definitive data was still lacking, that it would be folly to try to predict another massive quake. Weston was voicing the majority opinion. He’d done so with increasing support since the crisis had started.
Holleran went on. “From what I’ve read, triples aren’t all that unusual in intraplate settings like the one here. There were triples as recently as 1990 in the Sudan. The largest was a magnitude 7.3. The smallest a 6.7. They hit over a five-day period. In 1988, a rural area in Australia recorded three in the magnitude 6 range over a twelve-hour period.”
“We’ll have a panic on our hands if it leaks out we were even having this discussion,” Weston said. “We don’t need mass hysteria.”
“I’d say we already have it,” said Holleran. “How can we scare people any more than they already are?”