Bolivian quake deepens a mystery

Abstract

The 8.2 magnitude quake that recently struck beneath Bolivia has provided a sharp reminder to scientists that the Pacific plate's descent beneath South America is not following the textbook pattern.

When last week's magnitude 8.2 earthquake struck beneath Bolivia, days after a catastrophic shallower quake in Colombia, its effects were muted by the 640 kilometers of rock between the quake and the surface. But since then what may be the largest deep earthquake of the century has shaken up the scientific community, sending researchers scrambling to decode this signal from the planet's depths. And the first message they are taking from it is a sharp reminder that the descent of Pacific ocean crust into the mantle beneath South America isn't following the textbook pattern.

Deep earthquakes trace this process of subduction because they take place within the descending slab of tectonic plate, marking its slope to depths as great as 670 kilometers. The deepest ones are triggered, say recent theories (Science, 26 April 1991, p. 510), when the pressures of the deep Earth transform the rock from one crystalline phase to another. Thin layers of the new phase form and propagate so suddenly that the rock fails along these "anti-cracks," generating an earthquake.

In a textbook case, such quakes would be distributed evenly along the lowermost slab. But just days before the Bolivian quake, Emile Okal of Northwestern University in Evanston, Illinois, Stephen Kirby of the U.S. Geological Survey (USGS) in Menlo Park, California, Robert Engdahl of the USGS in Denver, and Wei-Chuang Huang of Northwestern had finished assembling convincing evidence that South American deep quakes don't follow this pattern. They recalculated the locations of this century's deep South American quakes using modern techniques and found that five of the seven biggest ones fell in two clusters, each cluster hundreds of kilometers away from the nearest smaller deep earthquakes. One of the clusters, a pair of quakes that struck in 1958 and 1967, sat in the middle of a 1000-kilometer-long gap in deep earthquake activity. And this latest quake fell right in the same lonely cluster of big, deep earthquakes, making it stand out all the more.

But why should the deep slab be especially prone to big earthquakes at that point? Seismologists Raymond Russo and Paul Silver of the Carnegie Institution of Washington's Department of Terrestrial Magnetism can't say for sure, but they have found indications that the slab may be under some unusual stresses right at the Bolivian cluster. Because of the westward movement of South America, the descending slab itself is steadily moving to the west, pulling through the ductile rock of the mantle like a garden hoe through soil. The mantle rock has to flow around the slab, and by studying how seismic waves from earthquakes propagate near the slab, Russo and Silver mapped this flow. They found that instead of being carried down with the descending slab, the mantle diverges like fine soil around a hoe, flowing north and south along the west face of the slab (Science, 25 February, p. 1105).

The flow isn't smooth, however; in three places it is disrupted where the slab changes its angle of descent. The Bolivian cluster of big deep earthquakes, Silver notes, lies at the bottom of one of these slab wrinkles. It's the most abrupt of the three, and because the mantle flow around the slab diverges just to the south of the wrinkle, the flow is most powerful in that region. No one is yet willing to speculate on just how the turbulent mantle flow might touch off giant earthquakes. But seismologists are hoping that this latest earthquake will help them sort out the complicated collision of continent, slab, and mantle beneath Bolivia.