Comprehensive Waveform Cross-correlation of Southern California Seismograms: Part 1. Refined Hypocenters Obtained Using the Double-difference Method and Tectonic Implications

We present preliminary results applying waveform cross-correlation to southern California seismograms for over 380,000 events between 1984 and 2002. Waveforms recorded by the SCSN are first extracted from the SCEDC data center in 50 s windows that include both P and S waves. The resulting online waveform archive uses about 0.5 TB on a RAID system. The traces are then re-sampled to a uniform 100 Hz sample rate and band-pass filtered to between 1 and 10 Hz. Next, we apply time domain waveform cross-correlation for P and S waves between each event and 100 neighboring events (identified from the catalog based on a 3-D velocity model of Hauksson (2000). We identify and save differential times from the peaks in the cross-correlation functions and use a spline interpolation method to achieve a nominal timing precision of 0.001 s. These differential times, together with existing P and S phase picks, are input to the double-difference program of Waldhauser and Ellsworth (2000). We define a grid across southern California and locate hypocenters near each grid node. Because some events may be located many times as hypocenters are calculated near successive grid-points, we assign a weight to each hypocenter and calculate a weighted average hypo-center for each earthquake. The new HypoDD hypocenters show improved clustering both horizontally and vertically, creating a more focused picture of the previously identified, spatially complex distributions of seismicity. In many cases, the late Quaternary faults, such as the Elsinore and Hollywood-Santa Monica faults appear to bracket the seismicity distributions; in other cases, the faults trace the median within a symmetric distribution of hypocenters. The depth distribution of the seismicity shows sudden changes across some of the major strike-slip faults, while regions of dip-slip faulting are often bound by dipping surfaces that are clearly defined by the deepest hypocenters. The seismicity around the southern San Andreas fault shows clear alignment along the Carrizo Plain segment while both the Mojave and Coachella Valley segments are dominated by off-fault hypocenters. A prominent horizontal boundary striking a few degrees north of west with a prominent depth change in the seismicity cuts across Banning Pass towards San Bernardino. Earthquake swarms in the Salton Sea at the south end of the San Andreas fault suggest the presence of two north-northwest striking seismic zones at the south end of the San Andreas fault. The seismicity along the San Jacinto fault forms sharp alignments that in most cases are adjacent to, but not coincident with, the mapped surface traces that are either parallel to the traces or form high angles to them. In the Los Angeles basin, the major aftershock sequences appear as densely focused clusters within a cloud of scattered background seismicity. The seismicity along the Newport-Inglewood fault forms a sharp alignment to the north and a diffuse distribution to the south, where the 1933 Long Beach earthquake occurred. Similarly, several clusters as well as scattered background seismicity extending from east to west across the basin illuminate the blind thrusts beneath the north edge of the basin. The major aftershock sequences such as 1992 Landers, 1994 Northridge, and 1999 Hector Mine form clusters, with distinct internal structures, illuminating secondary faults and a heterogeneous main fault rupture surface. Some of these alignments suggest that high angle cross-faults were activated by the mainshock.