RATE AND STATE OF BACKGROUND STRESS ESTIMATED FROM THE AFTERSHOCKS OFTHE 1989 LOMA-PRIETA, CALIFORNIA, EARTHQUAKE

Estimates of the tectonic stress state including loading rate and magn itude of background stress are derived from the spatial and temporal d istribution of Loma Prieta aftershocks, This technique was previously applied to the Landers aftershock sequence [Gross and Kisslinger, 1997 ] and is based upon the seismicity model of Dieterich [1994], Dieteric h's theory suggests that background seismicity should be proportional to stress rate and the number of aftershocks in an area should be prop ortional to the stress step experienced in that area. We used two inde pendently derived source models to compute the stress step from the ma inshock and to determine how effective that stress step was in trigger ing aftershocks. A background Stress state is then chosen which makes the stress steps at the aftershock locations most distinct from the st ress steps at hypocenters of background seismicity. The best fitting b ackground stress state has its greatest compressive stress plunging ap proximate to 17 degrees to N13 degrees E, and an intermediate stress v ery close in Magnitude to the least principal Stress, The small shear stresses at depth and low coefficient of friction suggest that high-pr essure pore fluids may be present inside active faults. The estimated stress rate of similar to 70 Pa/d (0.25 bar/yr) is comparable to the s tress rate found for the southern San Andreas Fault system, We found a best fitting ''effective'' coefficient of friction mu' approximate to 0.2 for the Loma Prieta area, significantly less than mu' approximate to 0.6 estimated for Landers in previous work, Variations in aftersho ck decay rate within the Loma Prieta aftershock zone are correlated wi th static stresses caused by postseismic slip as modeled by Burgmann e t al. [1997], Some of the postseismic slip occurred on structures that did not slip during the mainshock, so the postseismic and coseismic s tress step fields have different spatial distributions. The effectiven ess of slowly accumulated postseismic static stresses in triggering af tershocks is especially interesting because dynamic stresses are insig nificant in this case.