Bringing Seismicity into Focus: New Insights into Fault Behavior from Accurate Earthquake Locations

Bokelmann(1), G.H.R., Beroza(1), G., Cole(2), A., Ellsworth(2), W., Schaff(1), D., Waldhauser(2), F., Zanzerkia(1), E.

(1) Stanford University; (2) USGS, Menlo Park


Newly developed high-resolution images of seismicity on faults within the San Andreas system reveal structure that was previously obscured by earthquake location errors. By combining ordinary travel time picks with high-precision phase correlation measurements, improved earthquake locations reveal a wealth of detail previously hidden from view. In particular, hypocenters on the Hayward, Calaveras and San Andreas faults show that seismicity is highly organized in both space and time. Common features from all of these faults include: 1) a narrow (25-100 m) width of the fault core; 2) approximately horizontal streaks of hypocenters along the fault plane; 3) multiple recurrence of earthquakes of the same size at precisely the same spot on the fault (multiplets); and 4) regions devoid of earthquakes (holes). The narrow core of the fault is commonly embedded within a more diffuse volume of seismicity, which is itself often comprised of discrete narrow zones, i.e. faults. The streaks occur at multiple depths of the fault core at many locations, and have typical lengths of up to 5 km. Members of individual multiplets occur within a common hypocentroid, and must repeatedly rupture the same area of the fault. This suggests frictional or geometric controls on their location and size. The holes are perhaps the most intriguing feature, as in the long term, the fault must slip on these zones, either in earthquakes or by steady or episodic creep. The nucleation zones of the M 6.1 1966 Parkfield, M 6.2 1984 Morgan Hill and M 6.9 1989 Loma Prieta earthquakes all locate within holes near the base of the seismogenic zone, suggesting that these holes represent zones of locking. In contrast, the source zone of the 1992 slow earthquake at San Juan Bautista on the San Andreas fault (Linde, et al., 1996) appears to be coincident with shallow holes in the seismicity. If such holes represent likely nucleation regions of future earthquakes they suggest opportunities for focussed natural laboratory experiments in the context of the Plate Boundary Observatory.

Back to My Homepage , Stanford University or Department of Geophysics