抄録
Data from an array of proton magnetometers in central California indicate that a systematic decrease in magnetic field of about 2γ in 5 years has occurred in a localized region near Anzar, California, just north of the creeping section of the San Andreas fault. This field change has most likely resulted from changes in crustal stress in this region, although an unknown second-order effect of secular variation cannot be excluded as a alternate explanation. Tectonomagnetic models have been developed using dislocation modeling of slip on a finite section of fault. Assuming a fault geometry and rock magnetization, these models relate changes in stress, fault slip, and fault geometry to surface magnetic field anomalies. A large-scale anomaly, opposite in sense to that observed but of similar amplitude, would be expected to have accumulated in this area during the past 70 years. A localized 5-bar decrease in shear strain on the fault resulting from about 2cm of slip on a 0.25-km square patch at a depth of 1km beneath the surface trace of the fault opposite the magnetometer could explain the observed data and still be compatible with the geodetic strain measurements in the area. Other models of limited local slip are equally possible. The occurrence of a moderate magnitude earthquake in this region will allow comparison of stress changes estimated by different techniques.
