Results for stretching-dominated ridge models. The model is consistent with the large offset faults seen at the inside corners of slow-spreading segments, as well as with the asymmetry in magmatic accretion, because most magmatic accretion takes place on the side of the ridge with smaller fault offsets. more info

Scientists have long held the belief that the fracturing of the Earth's brittle outer shell into faults along the deep ocean's mountainous landscape occurs only during long periods when no magma has intruded. Challenging this predominant theory, findings from a completed study show how differences in mid-ocean ridge magma-induced activity produce distinctly different types of ocean floor faulting.

W. Roger Buck, Doherty Senior Research Scientist at the Lamont-Doherty Earth Observatory (LDEO), is one of a trio of scientists who developed these new models for faults seen at mid-ocean ridges where the Earth's tectonic plates split apart and basaltic magma rises to form the oceanic crust that today covers two-thirds of the planet. The scientists' work has culminated in the publishing of their findings in the April 7, 2005 issue of Nature.

Unlike faults on land, those formed along mid-ocean ridges are practically a dime a dozen. "The rate of fault generation across these ridges is a hundred times greater than on land," explains Buck. "And while land faults are easily eroded and often cut older faults in complex, hard-to-untangle ways, submarine faults break into newly formed crust and lithosphere and are little obscured by erosion. Recent observations show a huge range of fault types and sizes at ridges."

These combined factors make mid-ocean ridges "the place to learn about how faults form and grow." The team's findings challenge the standard view that all faults at these ridges result from tectonic stretching of thin near-ridge lithosphere (the Earth's brittle outer shell, where earthquakes are concentrated) in the absence of magma, hot molten rock from deep within the Earth. Among several recent observations that do not fit this standard model, two stand out: the first concerns where the faults form and the second deals with how far the faults slip. Faults formed at fast-spreading centers, like the East Pacific Rise, are tiny in comparison to faults that bound deep ocean hills at slow-spreading centers like the Mid-Atlantic Ridge. All ridge faults start off growing close to the ridge. Mid-Atlantic faults die only a short distance from where they are formed. In comparison, faults along the East Pacific Rise continue growing — although very slowly — much farther from the ridge axis. The new models show that these faults may form due to bending, not stretching, of the lithosphere.

Until a few years ago most scientists believed that the biggest faults at ridges account for around a kilometer of slip. But now we see some faults have slipped several tens of kilometers. Buck and his colleagues' study shows that special conditions may produce the larger offset "oceanic core complex" faults.

"Until the 1990s most people thought all slow-spreading crust was chopped up by many high-angle faults with the biggest of them having about a kilometer of offset," said Buck. "Then, a totally different kind of structure was found along parts of slow spreading ridges. At these oceanic core complexes hundreds of square kilometers of ocean floor are not cut by typical high-angle, ridge parallel faults and the magmatically accreted crust is thin or non-existent." One possible explanation is that these strange structures are related to faults that slip tens of kilometers and rotate so that they are nearly flat.

Modeling fault development is hard enough, but no group previously had combined simulation of fault development and magmatic dike intrusion, when magma flows and hardens into cross-cutting sheets in previously formed rock. "It is pretty clear that magma plays a big role in determining the style of faulting at a ridge. The places where the faults were smallest had the greatest supply of magma," said Buck. "At ridges, magma frequently cracks through the ridge axis." The team came up with a simple, yet very approximate, way to put dike intrusion into models of ridge faulting. The results of many experiments showed that different rates of intrusion could result in fundamentally different kinds of fault structures and topography, explaining the wide range of faults along the mid-ocean ridges. Long periods with no magma were not required.

"We were fairly constrained in some details while trying to simulate these fault formations. There are many possible ways that faults might form, and a lot of things we tried didn't work. Our models were based on the results of many experiments," said Buck. One problem they faced may have implications for how all faults form, including faults on land. "A big question is how the faults become weak: do they suddenly weaken when the rocks are stressed enough to break or is there slow wearing and smoothing of the fault as it slips?" The study shows that there has to be a sudden loss of some strength to make the kinds of small faults seen at fast-spreading ridges, but that much more weakening has to occur slowly with slip to develop faults with kilometers of offset seen at slow-spreading ridges.

Brian Tucholke, a scientist at Woods Hole Oceanographic Institution in Massachusetts, likes the idea that really big ridge faults form only when the amount of magmatism is "just right," calling this the "Goldilocks" effect. Tucholke has been studying ridge faulting for several decades and thinks he has evidence that supports the idea.

The team's study makes its debut at a time when questions about fault formation within the ocean's depths have captured global attention, due largely to the recent earthquakes near Sumatra, which caused the Indian Ocean tsunami. According to Buck, however, faulting across mid-ocean ridges is hardly a cause for alarm.

"There are frequent earthquakes caused by mid-ocean range faulting but they tend to be small," said Buck. "The critical difference lies in the fact that the Sumatra earthquake was caused by plates moving together as opposed to those moving apart as in the case of mid-ocean ridge faults. There isn't much to be afraid of, but there is a lot to be learned."