Upper-Mantle Anisotropy from SKS Splitting on a Dense Broadband Profile Across the Salton Trough, Southern California

Alex Kinsella, Shahar Barak, and Simon L. Klemperer Stanford University The Salton Trough is an active magma-dominated rift linking the Gulf of California to the San Andreas fault system. Because the rift is buried beneath a thick pile of Colorado River sedimentary rock, surprisingly little is currently known about the total volume of intrusion into the crust and the magma distribution within and beyond the rift margins. Our Broadband Salton Seismic Imaging Project, part of the larger USGS-NSF SSIP (Salton Seismic Imaging Project), deployed 42 broadband seismographs from 2011 to 2013 across the Salton Trough from San Diego to the Colorado River. Thirty-six of the stations were deployed along a WSW-ENE line across the trough, while the other six stations were deployed off-line to provide areal coverage. Our broadband study is intended to lead to a better understanding of magmatic dominated rifts. We analyzed shear wave splitting of SKS and SKKS phases from earthquakes with moment magnitude greater than 5.5 at distances 90° to 145°. Preliminary single-layer models based on the largest magnitude events show fast axes aligned roughly east-west at stations in the Peninsular Ranges west of the Salton Trough, as recognized by many previous studies, while in the center and east of the trough, the fast axes are roughly NNW-SSE (parallel to the axis of the Salton Trough). Delay times on the western side of the trough range from 0.5 to 1.5 seconds, but stations on the eastern side of the trough have generally shorter delay times, clustering between 0.5 and 0.8 seconds. These relatively large delay times indicate the mantle is the dominant source of the anisotropy. The W-E anisotropy to the west has previously been interpreted as due to either movement of the trailing edge of the Farallon plate over the last 20 million years or Cenozoic N-S compression in southern California. The NNW-SSE fast axes, within and east of the trough, might be explained by NW-SE mantle flow due to plate motion, basin-parallel fluid-filled cracks, or shearing associated with the San Andreas and related strike-slip faults. Some stations across our array have anomalous fast-axes, which could be caused by local heterogeneities, such as magma-filled cracks beneath the localized magmatic zones south of the Salton Sea. Additionally, there is some evidence for a much more complex model of anisotropy beneath the Salton Trough than we can model with our best-fit, single-horizontal-layer of mantle anisotropy. However, the seismically noisy environment of the trough makes it unlikely that we will have enough good events from sufficient back-azimuths, even by the end of our two-year deployment, to rigorously constrain multi-layer splitting models.