I was wondering if they'd ever get something out of the Parkfield science extravaganza. This latest tidbit comes from the 600 m deep seismometer array. Seems that there is a deep rumble that is sensitive to earth tides, and this has to be associated with pockets of fluid. The reasoning, of course, is that pockets of fluid can be extremely sensitive to slight changes in the local combined stress.
This pocket of fluid, trapped in rock can gather volumetric strain, and produce a significant effect at the narrow mouth. A classic effect is a Whistling Cave, where slight changes in air pressure cause strong winds at the entrance.
The rumbling at 15 km depth probably doesn't mean much to the real world of earthquakes.