Niu and Silver's Nature paper is a recent development in which they may actually see changes in the fault zone at seismogenic depths.

Migration of seismic scatterers associated with the 1993 Parkfield aseismic transient event

FENGLIN NIU, PAUL G. SILVER, ROBERT M. NADEAU & THOMAS V. MCEVILLY

Rice University, Carnegie Institution, UC Berkeley, and Lawrence Berkeley National Laboratory

The time-varying deformation field within a fault zone, particularly at depths where earthquakes occur, is important for understanding fault behaviour and its relation to earthquake occurrence. But detection of this temporal variation has been extremely difficult, although laboratory studies have long suggested that certain structural changes, such as the properties of crustal fractures, should be seismically detectable. Here we present evidence that such structural changes are indeed observable. In particular, we find a systematic temporal variation in the seismograms of repeat microearthquakes that occurred on the Parkfield segment of the San Andreas fault over the decade 1987–97. Our analysis reveals a change of the order of 10 m in the location of scatterers which plausibly lie within the fault zone at a depth of 3 km. The motion of the scatterers is coincident, in space and time, with the onset of a well documented aseismic transient (deformation event). We speculate that this structural change is the result of a stress-induced redistribution of fluids in fluid-filled fractures caused by the transient event.