Tuesday, January 03, 2006
Recognizing earthquakes in the geologic record
People are generally familiar with the relationship between earthquakes and faults – motion along a fault (like the San Andreas) causes an earthquake. However, motion along faults can occur without causing earthquakes (at least without causing large ones). Parts of the San Andreas, Hayward, and Calaveras faults in California all creep. They move slowly and steadily (perhaps through very small earthquakes) instead of moving through a series of jerks, thereby causing large earthquakes (like the 1906 Great San Francisco quake,the 1989 Loma Prieta quake, or the 1857 Fort Tejon event). So, not all faults cause earthquakes.
Ancient, inactive faults are very common, and a lot of geologists (like me, for example) study them since modern active faults can be hard to study. Large earthquakes generally occur at least 10 km below the earth's surface, and are therefore inaccessible (although the San Andreas Fault Observatory at Depth is an example of a project to drill through an active fault). These ancient faults are found in areas where rocks that were originally very deeply buried have been brought to the surface; for example in mountain ranges. So, these exhumed, ancient fault zones are great analogs for modern, active fault zones. If we want to understand the behavior of modern earthquake-generating (seismic) faults, we need to know which of these exhumed faults was seismic. In other words, we need a way to tell which of these faults were seismic, and which were creeping, and that's pretty tricky to do. The only generally agreed upon indicator of seismic motion along a fault is something called pseudotachylyte. Pseudotachylytes are glasses – they're formed when a fault slips fast enough that it generates enough heat to melt a part of the rocks adjacent to the fault. Therefore only seismogenic faults cause pseudotachylytes; creeping faults don't. The problem is that pseudotachylytes are very rare. Since they're so rare, papers describing them generate a fair bit of interest. In a recent paper in the journal Geology a team of geologists describe an unusually thick zone of pseudotachylytes along an exhumed subduction zone in Alaska. Subduction zone earthquakes are an area of intense research, because most of the world's largest earthquakes occur there. The tsunami-generating 2004 Sumatra-Andaman Islands earthquake occurred in a subduction zone, as did the 1964 Alaska earthquake. The pseudotachylyte described in the Geology paper may be a "fossil" of a great subduction zone earthquake. If this is true, this is as close as we can get to a great earthquake-generating fault without drilling through an active one, which is technologically and financially infeasible.
Ancient, inactive faults are very common, and a lot of geologists (like me, for example) study them since modern active faults can be hard to study. Large earthquakes generally occur at least 10 km below the earth's surface, and are therefore inaccessible (although the San Andreas Fault Observatory at Depth is an example of a project to drill through an active fault). These ancient faults are found in areas where rocks that were originally very deeply buried have been brought to the surface; for example in mountain ranges. So, these exhumed, ancient fault zones are great analogs for modern, active fault zones. If we want to understand the behavior of modern earthquake-generating (seismic) faults, we need to know which of these exhumed faults was seismic. In other words, we need a way to tell which of these faults were seismic, and which were creeping, and that's pretty tricky to do. The only generally agreed upon indicator of seismic motion along a fault is something called pseudotachylyte. Pseudotachylytes are glasses – they're formed when a fault slips fast enough that it generates enough heat to melt a part of the rocks adjacent to the fault. Therefore only seismogenic faults cause pseudotachylytes; creeping faults don't. The problem is that pseudotachylytes are very rare. Since they're so rare, papers describing them generate a fair bit of interest. In a recent paper in the journal Geology a team of geologists describe an unusually thick zone of pseudotachylytes along an exhumed subduction zone in Alaska. Subduction zone earthquakes are an area of intense research, because most of the world's largest earthquakes occur there. The tsunami-generating 2004 Sumatra-Andaman Islands earthquake occurred in a subduction zone, as did the 1964 Alaska earthquake. The pseudotachylyte described in the Geology paper may be a "fossil" of a great subduction zone earthquake. If this is true, this is as close as we can get to a great earthquake-generating fault without drilling through an active one, which is technologically and financially infeasible.
