Part I: Damage to the shallow Landers fault from the nearby Hector Mine earthquake Part II: The Spatial and Temporal Pattern of Shear Wave Splitting Along the Hector Mine Rupture Zone by John Vidale Department of Earth and Space Sciences University of California, Los Angeles 3-4pm Friday Feb 21, 2003 Refreshments served at 2:45pm Munk Conference Room Cecil and Ida M. Green Institute of Geophysics and Planetary Physics Scripps Institution of Oceanography University of California, San Diego http://mahi.ucsd.edu/seminar/ Abstract Part I: Damage to the shallow Landers fault from the nearby Hector Mine earthquake Evidence is accumulating that earthquake faults undergo mechanical damage that varies through the earthquake cycle. We have been watching the healing of damage on the shallow Johnson Valley fault after its rupture in the 1992 M7.4 Landers earthquake. The healing was interrupted by the 1999 M7.1 Hector Mine earthquake rupture 20-30 km away. The Hector Mine earthquake both strongly shook and permanently strained the Johnson Valley fault, adding damage discernible as a temporary reversal of the healing. The fault has since resumed the post-Landers trend of strength recovery. We speculate that fault damage by strong seismic waves helps explain earthquake clustering and seismicity triggering by shaking, and may play a role in friction reduction during faulting. Part II: The Spatial and Temporal Pattern of Shear Wave Splitting Along the Hector Mine Rupture Zone Extensive measurements of shear-wave splitting from aftershocks along the Hector Mine rupture zone show a broadly consistent fast direction, but also temporal and spatial variations in local stress directions. Aftershocks were recorded by several deployments in the year following the mainshock. A week-long 20-station, 100m array (Geom99) was deployed a few days after the mainshock. A second deployment with two arrays was in place November, 1999. NA99 was located near Geom99, 5km south of the mainshock and SEA99 was 10km farther south. In 2000, these two arrays were redeployed (NA00 and SEA00) and a third array (SWA00) was deployed on a fault strand 2km west of SEA00. Each array had a 20-station, 500-m line normal to the fault. The deployment of NA and SEA in 1999 and 2000 allows for examination of temporal changes in splitting. The wide distribution of the arrays allows for investigation of spatial variations. Rotation of the maximum compressive stress direction along fault strike is inferred from splitting measurements determined by cross-correlation technique. NA and SEA events were binned into three subsets based on local fault strike. The average measured splitting parameters indicate a 14.5 degree fast direction and a 33ms time lag. Rotation of the fast direction over one year was observed for the northern and central bins; however, sense of rotation varies along the fault. Also, measured time delays were lower in 2000 than 1999. Some spatial variation of splitting is observed. While splitting parameters measured at NA00 and SEA00 are fairly uniform, events recorded at SWA00 show very different fast directions and delay times. Delayed P and S arrivals indicate SWA00 is located atop low velocity material that appears to affect splitting measurements. Additionally, aftershocks within 0.25km of the fault have roughly fault-parallel fast directions. The rotation of fast azimuth with fault strike and time should help constrain the stress field evolution in the near-fault region following a major quake. We may be seeing stress regimes that were perturbed during the rupture relaxing over time to pre-event stress orientations.