Scientists Get a Closer Look at Mechanism of Deep Bolivian Quake

While big earthquakes typically wreack havoc for humankind, the biggest deep earthquake ever recorded - the magnitude 8.3 event that occurred 637 km beneath the surface of Bolivia on June 9, 1994 - has brought enlightenment - of sorts.

Instead of bringing death and destruction, the recent Bolivian quake, the largest quake of any type in recent decades, has given scientists one of the best probes yet of the Earth's interior. Within several minutes of the deep rupture, the quake benignly shook the ground from Argentina to Canada. But more important, the quake shook the Earth like a bell for months thereafter, giving scientists a raw data set of the planet's free oscillations - like none other.

The oscillations generated by the Bolivia quake are particularly valuable not only because they are from a big deep quake, but also because they were the first to be captured by a modern-day seismic network. Data from shallow events are not as useful for studying these normal mode vibrations because of the interference caused by surface seismic waves. By analyzing the frequency oscillations from deep events, scientists have been able determine the density and related properties of rock in various parts of the mantle. Geoscientists need these calculations to study the mechanism that shifts the Earth's plates. The Bolivia data set may be the most comprehensive - if not the most heavily scrutinized - yet, according to Steve Kirby of the U.S. Geological Survey in Menlo Park.

Of primary interest is elucidating the mechanism of deep-focus events - the nature of which, both seismologically and physically, has long been controversial. This topic was the focus of almost an entire day of discussion at AGU's Fall Meeting.

Even with the state-of-the-art seismic network, some answers may remain elusive. As Caltech seismologist Hiroo Kanamori puts it: "No matter what you use, there is some limit to the detail [of the seismological mechanism] that can be obtained."

Barriers to Understanding

The main obstacle to learning about the mechanism of deep-focus earthquakes has been that scientists have no direct way to observe the fault zone. In addition, until the Bolivian event, existing seismic and geodetic data had been somewhat limited. The first deep quake identified by scientists occurred in 1911. Prior to that time, scientists generally believed ruptures were not possible at such depths, even though deep quakes exhibit seismic energy patterns that are similar to those of shallow events.

Scientists have generally explained the similarities between the seismic radiation patterns of shallow and deep quakes by a "double couple" source model or various combinations thereof.

Scientists have thought for some time that conventional faults or brittle failure faulting cannot form in rock subjected to the conditions found at the depths of the Bolivian quake, which occurred where the Nazca plate is being subducted beneath the mantle of the South American continent. Deep quakes generally occur in areas where subduction is taking place.

At depths of 70 km and above on the other hand, where surface quakes are found, brittle-failure faulting is exactly what happens. Surface quakes occur in brittle crust that is simultaneously being pushed or pulled in more than one direction. When the rock eventually breaks, faults form, which in turn can generate quakes, as stress is built up along a given fault and then released.

Mechanisms Explored

Scientists have looked to other mechanisms to explain deep earthquakes from shear instability to delayed deformation. However, the problem has remained that there has been no sure-fire way to test these models. The Bolivia event presents the best shot at it for now.

One of the leading models has been transformational faulting, wherein a phase transition triggers the rupture. However, a large volume change is not thought to be generated because of the ambient shear stress. In this model, the atoms of the mantle rock known as olivine (Mg,Fe)₂SiO₄, are somehow rearranged to form a more condensed mineral called spinel, as both the temperature and pressure increase when the descending slab starts to move. Scientists believe that the differing strengths of the minerals may account for the sudden slip of the fault, as layers of spinel continue to descend into the Earth.

Perhaps not too surprisingly, reconciling the Bolivia data with the reigning model has not been an entirely straightforward process. Preliminary analyses reveal that the volume change in deep quakes is likely very, very small, while the stress drop is likely very large at about 1 kilobar. Yet the rupture velocity is very slow at about 1 km/s, especially with respect to the shear velocity, which is about 600 km/s. In other words, there is a lot of energy that is not being expended to produce seismic waves or, in effect is not accounted for, Caltech's Kanamori suggests, "Some other process must be involved."

To explain these data, Kanamori and his colleague of Masayuki KikuchiYokohama City University in Japan have proposed a new twist to the phase transition model: Melting may explain the conundrum, they assert, though acknowledging that this interpretation may be speculative at this time. Although melting would cause a volume increase, it would likely cancel out any volume reduction caused by the phase transition, which is reflected in the data.

Kanamori says the melting is likely confined to a small zone and doesn't involve a large volume. "But it does have a pronounced effect," Kanamori says.

Some scientists at the AGU Fall Meeting remained unconvinced of this interpretation. Nonetheless, Kanamori says, where the unaccounted for energy goes "is going to be a very important problem."

Determining the geometry of the fault plane is another way that scientists attempt to decipher the puzzle of deep quakes, in particular of the physical mechanism. Yet scientists are still debating whether the Bolivian data support or dismiss the phase transition model, among others

From the new data, scientists generally believe the rupture geometry of the Bolivian earthquake is practically horizontal. This geometry would make it rather different than the geometries determined for other deep quakes, where the fault plane usually is steeper than the auxiliary plane, Kanamori explains.

Yet some scientists say there are also arguments against the horizontal geometry existing at the depth of the earthquake. For one, the sense of slip for the Bolivia event points more to steepening of the slab than flattening, according to Paul Silver and his colleagues at the Carnegie Institution in Washington, D.C. For another thing, a recent tomographic study reveals that at depth, the slab is moving beneath the lower mantle at a 45° angle.

Either way, most scientists agree that various aspects the geometry of the subducting slab are uncertain. A more definitive answer should come when the slab is imaged from the seismic data - something scientists are working on, even though it may be a daunting task. Silver and his colleagues estimate the size of the subducted olivine wedge to be about 5 km thick.

Extraordinary Answers

For now, the consensus seems to be that a lot more analysis lies ahead. And one way or another, it seems that the answers will be deep beneath the Earth's surface.

References

Kirby, S.H., W.B. Durham, and L.A. Stern, Science 252, 216-225, 1991.

Return to Science and Society
Return to Starting Point