From Tiny Quakes, Major Insights into Earth

DURHAM, N.C. -- Duke University scientists are using waves from tiny earthquakes as geological equivalents of diagnostic X-rays. Emerging from the ground, the waves can reveal clues about the anatomy of a West Indian volcano. They can also help in finding geothermal steam in East Africa, and in zeroing-in on the epicenter of big earthquakes on the U.S. West Coast.

The "microearthquakes" involved are so small that they may not even register on some sensitive instruments, the researchers said. Nevertheless they offer useful probes of the earth's structure and geological processes because they occur much more frequently than do larger ones. They are also located at shallow enough depths for earthquake wave detectors called seismographs to be installed relatively nearby.

"We have these three major projects that are dovetailing together," said Peter Malin, a geology professor at the Nicholas School of the Environment and Earth Sciences. Malin's seismology group is now carrying out microearthquake studies on the volcanic island of Montserrat, in the Rift Valley in Kenya and along the notorious San Andreas earthquake fault in California.

The true goal of all three efforts is basic science, Malin said in an interview. "Our group is trying to understand the relationship between stresses in Earth's crust and the generation of these very small earthquakes," he said.

In the process, the scientists gain a fringe benefit -- information that is useful to society. The federal government, for example, has already committed more than $6 million through the National Science Foundation (NSF) to better understand the San Andreas fault, the locus of the great San Francisco earthquake as well as other big and damaging temblors.

The current focus of that project is a $1 million, 7,000-foot pilot drill hole near Parkfield, Calif. That is where Malin installed seismometers he helped design -- rugged enough to endure 220 degree Fahrenheit temperatures and pressures of 300 atmospheres -- that will help scientists decide where to finish drilling the San Andreas Fault Observatory at Depth (SAFOD).

Data from those seismometers have been transmitted by satellite since last year from the central California site to the Seismology Group's suite of offices at Duke. Such remote transmission enables Elyan Shalev, a Duke research scientist in seismology, to monitor for the latest microearthquakes day by day by clicking onto a computer.

"We are trying to study something that we can't get to," said Shalev. "No one has ever made measurements in a place where earthquakes actually happen. That's what we're trying to do in the San Andreas Fault."

Positioned 1‚ miles below the surface, the seismometers detect earthquakes occurring more than a half-mile deeper and 1.1 miles horizontally to the northeast of the drill hole's bottom, according to Shalev. By being located so far below ground, the seismometers can register much smaller and thus more frequent quakes than they would at the surface where conditions are much noisier, he said.

His average is three to four small quakes a day with magnitudes of 0.5 or less on a scale in which serious earthquakes register above 5. Small as they are, the vibrational signals these tiny quakes send out can give seismologists information about underground structures between the disturbances and the seismometers.

Some of the scientists in the SAFOD project will have to use those data to decide where to drill next, since the fault line is not actually directly underneath the pilot hole but rather east under a line of hills. Drilling in between will have to be angled both down and towards that range, where the target fault could easily be missed by a bore hole that is only six inches in diameter.

"The San Andreas Fault is not one large plane cutting through the earth," Shalev said. "Usually there are several branches, and their planes can be oriented at different angles." Even the best estimates of the fault's location might be off by 200 feet, he acknowledged. "It might be 200 feet deeper or shallower than we think, or further to the east or the west. It would be very easy to miss the whole thing."

Such locational uncertainties are why the geologists are paying so much attention to the information in every new microearthquake that finds its way to the data banks at Duke.

The hope is to have the observatory finished in time to record the next significant quake at Parkfield. That tiny ranching community dubs itself "the earthquake capitol of the world," because it is known to have repeated damaging temblors every 20 or 30 years -- frequent by earthquake standards. Scientists hope to learn why.

In Montserrat, a tropical island located in the Caribbean southeast of Puerto Rico, the question is not when the volcano will erupt but rather for how long. The Soufriere Hills Volcano's latest eruption began in 1995 and has since caused two-thirds of the population to flee.

This growing mountain continues to belch out not molten lava, but rather hot boulders that rumble down its slopes. By day, these big stones look dark like simmering coals, and by night they glow red. Even more perilous are the occasional explosions of hot gases that can rush across the hilly terrain faster than a car can drive. Such a "pyroclastic flow" killed a group of residents that attempted to farm land within the government-established danger zone.

"We're trying to understand the volcano system better, giving the government the information they need to know which areas are dangerous to occupy and which are not," said Shalev. Towards that end, a NSF-supported scientific team that includes Shalev and Malin has lowered four different kinds of instruments down four different 600 foot drill holes. Duke's contribution is a set of seismographs that might hint where molten magma is circulating underneath the summit.

Before losing pressure and transforming itself into hot solid rocks, as it does in this kind of volcano, the magma flows like a liquid through pipe-like conduits, the Duke scientists explained. The flowing magma, they hypothesized, might make the conduits reverberate like old steam pipes shaking in a Boston apartment. That vibration could cause tell-tale microearthquakes.

"We're trying to map the magma reservoir," said Shalev. "If we did, we could start to have some idea, for example, of how long this eruption will continue." Early returns from those instruments were not promising. Instead of picking up microearthquakes from the volcano's plumbing system, the seismographs seemed to be recording the larger vibrations of hot rocks falling from the summit.

"We picked up an event that was probably a magnitude 2 or 3, but we detected no small earthquakes," Malin said. "That's a surprise."

One explanation might be that the government authorities did not permit the scientists to lower their instruments very close to the volcano, Shalev added. Their fear was that the team that drilled the holes would not have had time to escape, even by helicopter, in the event of a pyroclastic flow.

The researchers' third major project will take place in Kenya's rift valley, where East Africa is splitting apart along a line of volcanoes. There, the geologists will use microearthquakes in attempts to track the location of underground fluids, in this case not hot magma but hot water. The goal is to determine where to drill for volcanically derived steam to fuel Kenya's next geothermal electric power plant, which would be the country's second.

"The good thing about geothermal is that it's a resource that is ours," said geophysicist Stephen Onacha, an employee of Kenya Power Generation Company (KenGen) who is at Duke getting his Ph.D. under Malin. "The big problem with geothermal is the money needed up front. One well costs about $1‚ million to drill.

"When you drill a well and you don't find steam, then $1.5 million is gone," Onacha added. "So our objective is to reduce the number of wells drilled. We need to put all possible detection methods together to learn the best places to look."

In 2000 Malin's group collaborated with Onacha's company on a U.S. Department of Energy-funded pilot study that focused on improving techniques for the combined use of microearthquakes as well as electromagnetic methods to map for potential underground geothermal steam reservoirs.

The new project will apply what was learned in 2000 at a new potential site at Longonot, Kenya and perhaps other sites. As in Montserrat, microearthquakes would signal the "rattling" of fluids, in this case water, moving through natural underground conduits. Electrical detection methods can also be used because water is a good conductor of electricity, Onacha said.

According to a study by the United Nations, which is funding the new project, without the additional information geothermal will "lose out" to fossil fuel plant development in meeting the nation's projected power demand.

Building diesel plants instead of geothermal ones would be "at odds with Kenya's recognized need and capacity to limit fossil fuel emissions and energy imports," the study said.