The shocking truth about electromagnetic phenomena associated with the 2004 M6.0 Parkfield earthquake

An 18 year long experiment to record resistivity and geoelectric potential changes prior to, during, and after a characteristic Parkfield earthquake revealed no measurable resistivity changes and only a barely detectable coseismic change in potential. The telluric array used in this experiment spanned an area of 5 km X 20 km straddling the rupture zone of the M6.0 earthquake and was capable of detecting resistivity changes of less than 1% along the fault. No changes in telluric transfer functions above 0.5% were associated with the earthquake. Coupled with an approximate 3-D resistivity model of the region, I can show that no resistivity changes larger than 0.4% could have occurred if we assume that the resistivity is perturbed by stress changes from a dislocation model of the rupture. This does not preclude some mechanism unrelated to the earthquake stress or strain, however. While no precursory electric potential change was observed, there was a step change on most telluric dipoles with a duration of ~3 hours that began at the time of the earthquake. Maximum changes of ~5 mV were seen on dipoles, as compared to naturally occurring telluric fluctuations of 400 mV or more. When resolved to individual electrodes, an electrode at the base of Carr Hill and approximately 200 m SW of the fault showed a maximum positive change of ~2.5mV while electrodes 1-2 km off the fault had negative changes of 0.5-1 mV. One simple model to explain a negative electrical potential gradient away from the fault is seismically triggered fluid flow moving away from the fault. However, two M5 aftershocks produced opposite polarity signals (fault was negative, and surroundings were positive).