Fault Creep

Creep cracks roads in distinctive patterns. Geology Guide photo

Fault creep is the name for the slow, constant slippage that can occur on some active faults without there being an earthquake. When people learn about it, they often wonder if fault creep can defuse future earthquakes, or make them smaller. The answer is "probably not," and this article explains why.

Terms of Creep

In geology, "creep" is used to describe any movement that involves a steady, gradual change in shape.

Soil creep is the name for the gentlest form of landsliding. Deformation creep takes place within mineral grains as rocks become warped and folded. And fault creep, also called aseismic creep, happens at the Earth's surface on a small fraction of faults.

Creeping behavior happens on all kinds of faults, but it's most obvious and easiest to visualize on strike-slip faults, which are vertical cracks whose opposite sides move sideways with respect to each other. Presumably it happens on the enormous subduction-related faults that give rise to the largest earthquakes, but we can't measure those underwater movements well enough yet to tell. The movement of creep, measured in millimeters per year, is slow and constant and ultimately arises from plate tectonics. Tectonic movements exert a force (stress) on the rocks, which respond with a change in shape (strain).

Strain and Force on Faults

Fault creep arises from the differences in strain behavior at different depths on a fault.

Down deep, the rocks on a fault are so hot and soft that the fault faces simply stretch past each other like taffy. That is, the rocks undergo ductile strain, which constantly relieves most of the tectonic stress. Above the ductile zone, rocks change from ductile to brittle. In the brittle zone, stress builds up as the rocks deform elastically, just as if they were giant blocks of rubber.

While this is happening, the sides of the fault are locked together. Earthquakes happen when brittle rocks release that elastic strain and snap back to their relaxed, unstrained state. (If you understand earthquakes as "elastic strain release in brittle rocks," you have the mind of a geophysicist.)

The next ingredient in this picture is the second force that holds the fault locked: pressure generated by the weight of the rocks. The greater this lithostatic pressure, the more strain that the fault can accumulate.

Creep in a Nutshell

Now we can make sense of fault creep: it happens near the surface where lithostatic pressure is low enough that the fault is not locked. Depending on the balance between locked and unlocked zones, the speed of creep can vary. Careful studies of fault creep, then, can give us hints of where locked zones lie below. From that, we may gain clues about how tectonic strain is building up along a fault, and maybe even win some insight into what kind of earthquakes may be coming.

Measuring creep is an intricate art because it occurs near the surface. The many strike-slip faults of California include several that are creeping. These include the Hayward fault in the east side of San Francisco Bay, the Calaveras fault just to the south, the creeping segment of the San Andreas fault in central California, and part of the Garlock fault in southern California.

(However, creeping faults are generally rare.) Measurements are made by repeated surveys along lines of permanent marks, which may be as simple as a row of nails in a street pavement or as elaborate as creepmeters emplaced in tunnels. At most locations, creep surges whenever moisture from storms penetrates into the soil—in California that means the winter rainy season.

Creep's Effect on Earthquakes

On the Hayward fault, creep rates are no greater than a few millimeters per year. Even the maximum is just a fraction of the total tectonic movement, and the shallow zones that creep would never collect much strain energy in the first place. Creeping zones there are overwhelmingly outweighed by the size of the locked zone. So if an earthquake that might be expected around every 200 years, on average, occurs a few years later because creep relieves a bit of strain, no one could tell.

The creeping segment of the San Andreas fault is unusual. No large earthquakes have ever been recorded on it. It's a part of the fault, about 150 kilometers long, that creeps at around 28 millimeters per year and appears to have only small locked zones if any. Why is a scientific puzzle. Researchers are looking at other factors that may be lubricating the fault here. One factor may be the presence of abundant clay or serpentinite rock along the fault zone. Another factor may be underground water trapped in sediment pores. And just to make things a little more complex, it may be that creep is a temporary thing, limited in time to the early part of the earthquake cycle. Although researchers have long thought that the creeping section may stop large ruptures from spreading across it, recent studies have cast that into doubt.

The SAFOD drilling project succeeded in sampling the rock right on the San Andreas fault in its creeping section, at a depth of almost 3 kilometers. When the cores were first unveiled, the presence of serpentinite was obvious. But in the lab, high-pressure tests of the core material showed that it was very weak because of the presence of a clay mineral called saponite. Saponite forms where serpentinite meets and reacts with ordinary sedimentary rocks. And clay is very effective at trapping pore water. So, as often happens in Earth science, everyone seems to be right.